development of a solar pv energy assessment tool for eg-audit ltd

DEVELOPMENT OF A SOLAR PV ENERGY ASSESSMENT TOOL FOR EG-AUDIT LTD.

PROJECT REPORT

By DANIEL RAYMOND OWEN*

Department of Chemical & Biological Engineering, the University of Sheffield

Supervised by:

Dr. Alan Dunbar & Kevin Aylward

28th April 2016

*Author for correspondence

Acknowledgements

I wish to thank Dr. Alan Dunbar, Director of Student Support & Senior Lecturer in Energy, and Kevin

Aylward, co-founder & chief consultant for EG-Audit, who were instrumental in the successful

completion of this project. Kevin Aylward was solely responsible for providing the electricity usage

data for analysis and helped shed light on the auditing and billing industry. I’d like to especially

thank Dr. Alan Dunbar for his full co-operation and support; and helping to initiate and define the

scope of this project.

This research was supported by the EG-Audit & the University of Sheffield; however the contents do

not necessarily reflect their views and policies. Any mention of external businesses or commercial

products does not constitute in their endorsement or recommendation for use.

Development of a Solar PV Energy Assessment Tool for EG-Audit Ltd.

ii

Abstract

Solar PV is growing increasingly essential to the potential delivery of a future energy system that can

successfully overcome the energy trilemma facing current governments and industries. The UK’s

unstable microgeneration policies are considered one of the principal obstacles to developing

progressively sustainable energy infrastructure. This project will investigate the potential effect of

recent changes to FiT and export tariff rates on the financial incentive of PV systems and speculate

the impact this may have on the PV industry as a whole.

Analysis of currently commercially available solar assessment tools showed a significant variance in

financial results (follow Link 3 below). The delivery of inconsistent predictions may create a feeling

of ambiguity/mistrust in consumers, diminishing future PV installation rates. Moreover, some tools

were found to omit important solar irradiance determining system specification factors. Developing

a wholly transparent assessment tool, displaying all assumptions used during quote generation

whilst maintaining a high level of accuracy, is the key aim of this project (follow Link 1 below). This

report outlines the research completed, and methods employed, for the development of said tool.

Alternative solar calculators consistently employ a 50% export ratio assumption for financial

assessment calculations; in reality, export levels are heavily dependent upon electricity usage

behaviours. Data supplied by EG-Audit Ltd. will allow the application of identified usage behaviours

for six case study properties during quote generation (follow Link 4 below). The ability to make

usage behaviour change recommendations derived from the results of this study. Proposals for PV

system optimisation were also facilitated by this unique assessment tool feature.

Upon the introduction of the February 2016 microgeneration tariff rates, the total proportion of

lifetime profits that FiT payments account for falls from 56.3% to 30.9%; with expected return on

investment, for the system modelled in this report, falling from 4.8% to 1.2%.

Of the assessed case studies, the export ratio was found to range from 5.5% to 21.6% for the six case

studies; well below the 50% assumption employed by the alternative calculators. The developed

tool predicted an average lifetime profit of £7,940 versus an average profit of £6,760 predicted by

the alternative calculators. In order to determine the developed tool’s accuracy, EG now need to

compare the results of a completed assessment against meter readings after PV installation.

1. EG-Audit Solar Assessment Tool tinyurl.com/gplerwm

2. Assessment Tool User Guide tinyurl.com/gp5bjmr

3. Alternative Calculator Result Analysis tinyurl.com/hyr65w5

4. Case Study Analysis tinyurl.com/hyr65w5


iii

Table of Contents

ABSTRACT II

TABLE OF CONTENTS III

LIST OF FIGURES VI

LIST OF TABLES IX

NOMENCLATURE & ABBREVIATIONS XI

1.0 LITERATURE REVIEW 1

1.1 PROJECT MOTIVATION 1

1.2 OBJECTIVES 3

1.3 INTRODUCTION 4

1.3.1 EG-Audit 4

1.3.2 Microgeneration Certification Scheme 4

1.3.3 Solar Power 5

1.3.4 Solar Thermal Energy 5

1.3.5 Photovoltaic Solar Energy 7

1.3.6 PV system equipment 12

1.3.7 Alternative PV options 12

1.3.8 How does Silicon PV work? 15

1.3.9 Bandgap 16

1.3.10 Recombination 18

1.4 ELECTRICITY METERING 19

1.4.1 Metering Point Administration Number 19

1.4.2 Half Hourly Data 21

1.4.3 Property Types 22

1.5 SOLAR PV GENERATION FACTORS 24

1.5.1 Insolation/Irradiance 24

1.5.2 Seasonal Variation 26

1.5.1 Sun Path 27

1.5.2 Pollution Levels 29

1.5.1 Panel Inclination & Orientation 31

1.5.2 Shading 34

1.5.3 Weather Patterns 36

1.5.4 Degradation 37

1.5.5 Equipment Losses 37

1.6 LIFE CYCLE OF PV MODULES 39

1.6.1 Energy Pay-Back Time 39


iv

1.6.2 Cups of Tea Equivalent 40

1.6.3 Greenhouse Gas Equivalence 41

1.6.4 Offset CO2 equivalent calculations 42

1.7 FINANCIAL ASSESSMENT 43

1.7.1 Self-Consumption 43

1.7.2 Export Tariff 45

1.7.3 Feed-in Tariff 46

1.7.4 Annual Savings 47

1.7.5 Profit, PBP, & Return on Investment 47

1.8 ALTERNATIVE CALCULATORS 50

1.8.1 Calculators Selected 50

1.8.2 Review Method 51

2.0 RESULTS AND ANALYSIS 53

2.1 ANALYSIS OF ALTERNATIVE CALCULATORS 53

2.1.1 Review of Input Factors 53

2.1.2 Review of Results Presentation 55

2.1.1 Numerical Results Comparison 59

2.2 ASSESSMENT TOOL RESULTS COMPARISON 62

2.2.1 Generation Profile 62

2.2.2 Actual Export Ratio 63

2.2.3 Results Comparison 66

2.3 ELECTRICITY USAGE TRENDS 69

2.3.1 Annual Usage Trends 69

2.3.1 Weekly Usage Trends 71

2.3.2 Daily Usage Trends 72

2.4 CASE STUDY RESULTS 73

2.4.1 Commercial Property Usage Trends 73

2.4.2 Educational Property Usage Trends 75

2.4.1 Industrial Property Usage Trends 77

2.4.2 Light Commercial Property Usage Trends 79

2.4.3 Residential Property Usage Trends 82

2.4.4 Social Property Usage Trends 84

2.4.5 Case Study Summary 87

3.0 CONCLUSIONS 88

4.0 FUTURE DEVELOPMENTS 90

5.0 REFERENCES 92


v

6.0 APPENDICES 100

6.1 SOLAR PV GENERATION APPENDICES 100

6.2 LIFE CYCLE APPENDICES 101

6.3 RAW DATA FOR TOOL 102

6.4 ALTERNATIVE CALCULATORS GRAPHS 108

6.5 CASE STUDY RESULTS PRINT-OUTS 110

6.5.1 Commercial Property 110

6.5.2 Educational Property 111

6.5.3 Industrial Property 112

6.5.4 Light Commercial 113

6.5.5 Residential 114

6.5.6 Social 115


vi

List of Figures

FIGURE 1.3-1: EVACUATED TUBE SOLAR HEATING AND COOLING SYSTEM (EQUILIBRE PERSONNEL, 2016). 6

FIGURE 1.3-2: PARABOLIC TROUGH CONCENTRATED SOLAR POWER PLANT NEAR IN CALIFORNIA (SKYFUEL,

2011). 6

FIGURE 1.3-3: 2010 PROJECTED TO 2015 EU PV MARKET SHARE (ADAPTED FROM VATANSEVER, ET AL., 2012). 7

FIGURE 1.3-4: EXAMPLE PV SYSTEMS FROM SMALL TO UTILITY SCALE (FROM LEFT TO RIGHT: ROOF TOP ≈ 4

KW, FIELD ARRAY ≈ 1-5 MW, SOLAR POWER STATION < 1 GW) 8

FIGURE 1.3-5: COLAS ROUTE SOLAIRE WATTWAY (GUERRINI, 2016). 8

FIGURE 1.3-6: SOUTH KOREA’S SOLAR BIKE LANE SHADING SYSTEM (ALTER, 2015). 9

FIGURE 1.3-7: SWANSON’S LAW OF EXPERIENCE ANALYSIS APPLIED TO PV MODULE COST PER WATT (ADAPTED

FROM SEMI, 2014). 9

FIGURE 1.3-8: PROJECTED GLOBAL CUMULATIVE CAPACITY UP TO 2019 (ADAPTED FROM EPIA, 2014). 10

FIGURE 1.3-9: GLOBAL GRID PARITY BREAKDOWN AS OF 2014 (ADAPTED FROM SHAH, ET AL., 2014). 11

FIGURE 1.3-10: GLOBAL MARKET SHARE BY PV TECHNOLOGY FROM 1990 TO 2013 (ADAPTED FROM

FRAUNHOFER INSTITUTE FOR SOLAR ENERGY SYSTEMS, 2015). 13

FIGURE 1.3-11: NREL MULTI-STRAND BEST RESEARCH CELL EFFICIENCIES CHART (NREL, 2015). 14

FIGURE 1.3-12: GENERAL ARRANGEMENT OF SILICON PV CELL. 15

FIGURE 1.3-13: ABSORPTION COEFFICIENT OF SILICON AGAINST WAVELENGTH ON A LOGARITHMIC SCALE

(DATA COLLECTED BY AND ADAPTED FROM GREEN, 1995). 17

FIGURE 1.3-14: BANDGAP AND ASSOCIATED MAXIMUM EFFICIENCY FOR SINGLE LAYER SILICON P-N JUNCTION

AS OUTLINED BY AND ADAPTED FROM SHOCKLEY AND QUEISSER, 1961. 17

FIGURE 1.4-1: TYPICAL PRINTED FORMAT FOR MPAN NUMBER INCLUDED ON MOST ELECTRICITY BILLS. 19

FIGURE 1.5-1: YEARLY TOTAL IRRADIATION IN KWH/M2 FOR UK (AVERAGING PERIOD 1993-2007,

INFORMATION COURTESY OF THE MET OFFICE). 24

FIGURE 1.5-2: MCS POSTCODE SPECIFIED REGIONS FOR WHICH IRRADIANCE VALUES HAVE BEEN AVERAGED

ACCORDING TO MET. OFFICE DATA COLLECTED FROM 1993-2007 (ELECTRICAL CONTRACTORS

ASSOCIATION, 2012). 25

FIGURE 1.5-3: MONTHLY GENERATION AS A PROPORTION OF ANNUAL GENERATION FOR THE UK; ADAPTED

FROM HISTORICAL DATA GATHERED BY ELECTRICAL CONTRACTORS ASSOCIATION (2012). 26

FIGURE 1.5-4: EOT VERSUS D VALUES USED TO MODEL THE PATH OF THE SUN FOR GENERATION

CALCULATIONS. 28

FIGURE 1.5-5: SEASONAL VARIANCE OF SUNRISE AND SUNSET TIMES FOR CH2 4DW (+53.5˚, +2.6˚). 29

FIGURE 1.5-6: SOLAR RADIATION SPECTRUM FOR AM1.0, HIGHLIGHTING ABSORPTION BANDS FOR VARIOUS

ATMOSPHERIC GASES (ADAPTED FROM RHODE, 2016). 30

FIGURE 1.5-7: OPTIMUM INCLINATION ANGLE ACROSS THE UK (ADAPTED FROM MIDSUMMER ENERGY, 2016).

32

FIGURE 1.5-8: VARIABLES DEFINITION SOLAR RADIATION EQUATIONS ON A TILTED SURFACE. 33


vii

FIGURE 1.5-9: DIFFERENT TRACKING SYSTEMS EMPLOYED BY FREE MOUNTED PV SOLAR ARRAYS (ADAPTED

FROM SEDONA SOLAR TECHNOLOGY, 2016). 33

FIGURE 1.5-10: EFFECT OF ARRAY ORIENTATION AND INCLINATION ON SYSTEM PERFORMANCE FOR A

PROPERTY IN THE MIDDLE OF UK. VALUES ARE REPRESENTATIVE OF A PERCENTAGE OF THE MAXIMUM

YIELD (SOUTH FACING, 35˚ INCLINATION). ALL VALUES ARE TAKEN FROM ELECTRICAL CONTRACTORS

ASSOCIATION (2012). 34

FIGURE 1.5-11: DIFFUSE/GLOBAL IRRADIANCE RATIO PLOT FOR UK LATITUDE VALUES TAKEN FROM MUNEER

(2004). 35

FIGURE 1.6-1: EPBT FOR PROCESSES INVOLVED IN THE MANUFACTURE, INSTALLATION, AND OPERATION OF A

CRYSTALLINE SI SOLAR MODULE INSTALLED IN SYDNEY IN 2000 AND ESTIMATED VALUES IN 2010

(ADAPTED FROM BLAKERS & WEBER, 2000). 40

FIGURE 1.6-2: CO2 EQUIVALENT EMISSION LEVELS FOR DIFFERENT ENERGY SOURCES CURRENTLY, INCLUDING

FORECASTED CCS LEVELS (SOLAR INSOLATION = 1700 KWH/M2 ASSUMED) (ADAPTED FROM WORKING

GROUP III, 2014). 41

FIGURE 1.7-1: PES AREAS OF THE UK AND THEIR DISTRIBUTOR CODE NUMBER (ADAPTED FROM OFGEM, 2016).

43

FIGURE 1.7-2: AVERAGE CAPITAL COST FOR INCREASING INSTALLATION SIZE IN UK (ADAPTED FROM THE

GREENAGE, 2016) 48

FIGURE 1.8-1: PROPERTY SELECTED AS ‘STANDARD’ TO COMPARE ASSESSMENT RESULTS FROM ALTERNATIVE

CALCULATORS. ROOF SELECTED FOR PV INSTALLATION HAS BEEN HIGHLIGHTED (ADAPTED FROM

GOOGLE MAPS). 51

FIGURE 2.1-1: LAYOUT OF EG’S SOLAR ASSESSMENT INPUT PAGE. 54

FIGURE 2.1-2: NUMBER OF RESULTS DISPLAYED BY RIVAL CALCULATORS INCLUDING DISCERNIBLE

INFORMATION FROM GRAPHS (SEE SECTION 6.4), RANKED IN DESCENDING LEVEL OF DETAIL. 56

FIGURE 2.1-3: NOVEL RESULT PRESENTATION STYLES (IMAGES TAKEN FROM EG’S ASSESSMENT TOOL PRINT-

OUTS). 58

FIGURE 2.1-4: PROFITS PREDICTED BY RIVAL CALCULATORS RANKED IN ASCENDING ORDER. 60

FIGURE 2.1-5: PROSPECTIVE SAVINGS BREAKDOWN BEFORE AND AFTER CHANGES TO THE FIT RATE (ASSUMED

12.2P/KWH BEFORE JANUARY 2016 AND 4.4P/KWH AFTER FEBRUARY 2016). 61

FIGURE 2.2-1: AVERAGE SEASONAL GENERATION TREND FOR SYSTEM SPECIFIED IN TABLE 1.8-1, DATA IS

TAKEN FROM EG’S FINAL ASSESSMENT TOOL. 62

FIGURE 2.2-2: AVERAGE DAILY USAGE VERSUS GENERATION CURVES FOR THE COMMERCIAL AND RESIDENTIAL

PROPERTIES FOR A 4KWP SOLAR INSTALLATION AND ELECTRICITY USAGE OF 3,800KWH/ANNUM. 63

FIGURE 2.2-3: SEASONAL EXPORT VALUES AD A PROPORTION OF TOTAL ANNUAL EXPORT, AVERAGED FOR ALL

CASE STUDY DATA SUPPLIED BY EG. 65

FIGURE 2.2-4: ENVIRONMENTAL RESULTS FOR 4KWP SYSTEM DEFINED IN TABLE 1.8-1. 67

FIGURE 2.3-1: ANNUAL USAGE TRENDS FOR ALL SIX CASE STUDIES SUPPLIED BY EG. 69


viii

FIGURE 2.3-2: AVERAGE SEASONAL USAGE AS A PROPORTION OF TOTAL ANNUAL USAGE FOR ALL PROPERTIES

IN THE SAMPLE ANALYSED FROM HALF-HOURLY DATA. 70

FIGURE 2.3-3: WEEKLY USAGE TRENDS FOR ALL SIX PROPERTY TYPES FROM DATA SUPPLIED BY EG. 71

FIGURE 2.3-4: AVERAGE USAGE AS A PROPORTION OF TOTAL WEEKLY USAGE FOR ALL PROPERTIES. 71

FIGURE 2.3-5: DAILY USAGE TRENDS FOR ALL SIX PROPERTY TYPES FROM DATA SUPPLIED BY EG. 72

FIGURE 2.3-6: AVERAGE USAGE AS A PROPORTION OF TOTAL DAILY USAGE FOR ALL PROPERTIES. 72

FIGURE 2.4-1: AVERAGE WEEKLY USAGE TREND FOR COMMERCIAL PROPERTY. 74

FIGURE 2.4-2: AVERAGE DAILY USAGE TREND FOR COMMERCIAL PROPERTY. 74

FIGURE 2.4-3: AVERAGE WEEKLY USAGE TREND FOR EDUCATIONAL PROPERTY. 76

FIGURE 2.4-4: AVERAGE DAILY USAGE TREND FOR EDUCATIONAL PROPERTY. 76

FIGURE 2.4-5: AVERAGE WEEKLY USAGE TREND FOR INDUSTRIAL PROPERTY. 78

FIGURE 2.4-6: AVERAGE DAILY USAGE TREND FOR INDUSTRIAL PROPERTY. 78

FIGURE 2.4-7: AVERAGE WEEKLY USAGE TREND FOR LIGHT COMMERCIAL PROPERTY. 80

FIGURE 2.4-8: AVERAGE DAILY USAGE TREND FOR LIGHT COMMERCIAL PROPERTY. 81

FIGURE 2.4-9: AVERAGE WEEKLY USAGE TREND FOR RESIDENTIAL PROPERTY. 83

FIGURE 2.4-10: AVERAGE DAILY USAGE TREND FOR RESIDENTIAL PROPERTY. 83

FIGURE 2.4-11: AVERAGE WEEKLY USAGE TREND FOR SOCIAL PROPERTY. 85

FIGURE 2.4-12: AVERAGE DAILY USAGE TREND FOR SOCIAL PROPERTY. 86

FIGURE 6.1-1: REFLECTIVITY OF A POLISHED SILICON WAFER FOR THE MAJORITY OF THE WAVELENGTHS OF

THE VISIBLE SPECTRUM (GREEN M. A., 1995). 100

FIGURE 6.2-1: MANUFACTURING PROCESS FOR SILICON BASED PV MODULES (PENG, LU, & YANG, 2013). 101

FIGURE 6.3-1: EXAMPLE OF HALF HOURLY DATA SUPPLIED BY EG. ALL VALUES ARE GIVEN IN KWH/HALF HOUR.

102

FIGURE 6.3-2: EXAMPLE OF MSC SPECIFIED IRRADIANCE CHART FOR ZONE 1 (ELECTRICAL CONTRACTORS

ASSOCIATION, 2012). 106

FIGURE 6.4-1: FINANCIAL RESULTS GRAPH TAKEN FROM THE ENERGY SAVING TRUST (2016). 108

FIGURE 6.4-2: GENERATION AND FINANCIAL GRAPHS TAKEN FROM SOLAR GUIDE (2015). 108

FIGURE 6.4-3: FINANCIAL RESULTS GRAPH TAKEN FROM SOLAR WORLD UK (SOLAR WORLD UK, 2016). 109

FIGURE 6.4-4: FINANCIAL RESULTS GRAPH TAKEN FROM SOLAR PANELS UK (SOLAR PANELS UK, 2016). 109


ix

List of Tables

TABLE 1.3-1: SITUATIONAL ANALYSIS FOR DIFFERENT PHOTON ENERGY LEVELS. 16

TABLE 1.4-1: PROFILE CLASS DEFINITIONS. 19

TABLE 1.4-2: UNIT CATEGORIES AND THEIR AVERAGE UK DEFINED PERIODS OF USE. 20

TABLE 1.4-3: MTC VALUES AND THEIR DEFINITIONS (ENERGY LINX, 2016). 20

TABLE 1.4-4: PRICING STRUCTURE UNIT PRICES USED FOR TARIFF SUGGESTION ASPECT OF TOOL (EG-AUDIT,

2016). 21

TABLE 1.4-5: PROPERTY CLASSIFICATIONS AGREED WITH EG FOR ANALYSIS IN THE ASSESSMENT TOOL. 22

TABLE 1.5-1: THE EFFECTS OF VARYING HOT-SPOT TEMPERATURES (ABDALLA, 2013) 35

TABLE 1.5-2: CLOUD COVERAGE DATA FOR TWO OPPOSITE BOUNDARY AREAS WITHIN ZONE 13 (NASA, 2016).

36

TABLE 1.5-3: LOSSES AND THEIR TYPICAL MAGNITUDES IN A PV SYSTEM. 37

TABLE 1.7-1: DOMESTIC COST PER UNIT CONSUMED FOR EACH DNO AS OF JANUARY 2016 (GARDNER, 2015;

EG-AUDIT, 2016). 44

TABLE 1.7-2: UK ‘FLOOR PRICE’ EXPORT TARIFF FOR SOLAR INSTALLATIONS SINCE INTRODUCTION (FEED-IN

TARIFFS, 2016). 45

TABLE 1.7-3: UK FIT RATE FOR SOLAR INSTALLATIONS SINCE INTRODUCTION (FEED-IN TARIFFS, 2016). 46

TABLE 1.7-4: INVESTMENTS OF COMPARABLE MAGNITUDE TO SOLAR PV CAPITAL OUTLAY AND THEIR

AVERAGE ROIS. 49

TABLE 1.8-1: UK AVERAGE VALUES USED TO GENERATE QUOTES FOR ALTERNATIVE CALCULATOR. 52

TABLE 2.1-1: RIVAL SOLAR CALCULATORS RANKED FOR FACTORS EMPLOYED WHEN GENERATING QUOTES. 53

TABLE 2.1-2: NOVEL FEATURES INCLUDED IN EG’S ASSESSMENT TOOL. 55

TABLE 2.1-3: RIVAL SOLAR CALCULATORS RANKED FOR DISPLAYED RESULTS FOR GENERATED QUOTES. 56

TABLE 2.1-4: NOVEL RESULTS PRESENTED IN EG’S ASSESSMENT TOOL. 58

TABLE 2.1-5: ALL NUMERICAL RESULTS GAINED FROM ALTERNATIVE CALCULATOR REVIEW. 59

TABLE 2.2-1: EXPORT RATIO VARIANCE FOR EG SUPPLIED CASE STUDIES. 64

TABLE 2.2-2: SEASONAL EXPORT AS A PROPORTION OF TOTAL ANNUAL EXPORT FOR EACH CASE STUDY. 65

TABLE 2.2-3: ADDITIONAL INPUTS AND THE VALUES SELECTED FOR RESULT RETRIEVAL FROM FINAL

ASSESSMENT TOOL. 66

TABLE 2.2-4: ALL RESULTS FOR DIFFERENT PROPERTY TYPES RELATIVE TO ALTERNATIVE CALCULATOR RESULTS.

66

TABLE 2.4-1: TARIFF COST BREAKDOWN FOR COMMERCIAL PROPERTY. 73

TABLE 2.4-2: TARIFF COST BREAKDOWN FOR EDUCATIONAL PROPERTY. 75

TABLE 2.4-3: TARIFF COST BREAKDOWN FOR INDUSTRIAL PROPERTY. 77

TABLE 2.4-4: TARIFF COST BREAKDOWN FOR LIGHT COMMERCIAL PROPERTY. 80

TABLE 2.4-5: TARIFF COST BREAKDOWN FOR RESIDENTIAL PROPERTY. 82

TABLE 2.4-6: TARIFF COST BREAKDOWN FOR SOCIAL PROPERTY. 85

TABLE 2.4-7: SUMMARY OF CASE STUDY KEY RESULTS DISCUSSED IN SECTION 2.4. 87


x

TABLE 6.3-1: POSTCODE ZONING DATA (ELECTRICAL CONTRACTORS ASSOCIATION, 2012). 103

TABLE 6.3-2: AVERAGE LATITUDE AND LONGITUDES FROM PROPERTIES WITHIN THE DESIGNATED MCS ZONES.

THESE VALUES WERE USING TO PLOT THE PATH OF THE SUN THROUGHOUT THE YEAR IN THE

ASSESSMENT TOOL. 105

TABLE 6.3-3: DIFFUSE/GLOBAL IRRADIATION RATIOS USED TO CALCULATE LOSS OF INCIDENT ENERGY AT

DIFFERENT SHADING LEVELS (MUNEER, 2004). 107

TABLE 6.3-4: PRICING STRUCTURE PRICES P/KWH AND QUARTER METER CHARGES USED TO CALCULATE

PRICING STRUCTURE COMPARISONS. 107


xi

Nomenclature & Abbreviations

Acronym Definition

AM Air Mass Coefficient

CAT Centre for Alternative Technologies

CSP Concentrated Solar Power

DNO Distribution Network Operator

EG EG-Audit Ltd

EoT Equation of Time

EPBT Energy Pay Back Time

E/W Evening and Weekend

FiT Feed-in Tariff

NREL National Renewable Energy Laboratory

LCOE Levelised cost of Electricity

LLF Line Loss Factor

LSTM Local Standard Time Meridian

MCS Microgeneration Certification Scheme

MPAN Metering Point Administration Number

MTC Meter Time Switch Code

PBP Payback Period

PES Public Electricity Supplier

PV Photovoltaic

ROI Return on Investment

STC Standard Test Conditions

STE Solar Thermal Energy

STOD Seasonal Time of Day

SRH Shockley-Read-Hall

TC Time Correction

W/N/E/W Weekday/Night/Evening/Weekend

1.0 Literature Review

1.1 Project Motivation

In recent times, research into alternative, renewable energy sources has arguably been one of the

most discussed topics in the science community. Much data has been gathered researching the

potential of the different technologies available, with solar energy regularly touted as the most

“essential component of any serious strategy to mitigate climate change” (Energy Initiative

Massachusetts Institute of Technology, 2015).

For the last two decades, photovoltaics (PV) have been the fastest growing renewable energy

industry in respect to their present size (Hoffman, 2006). Continuing at the present growth rate of

40% per annum will allow PV to become the world’s largest energy source by 2040 (Honsberg &

Bowden, 2015). Solar panel companies are quick to promote the financial advantages of installing

their equipment to customers; yet early adopters of the technology are sometimes led to believe

that they can earn hundreds of thousands of pounds, when the reality can be far less optimistic

(Lonsdale, 2015). Companies would also be at risk of losing their Microgeneration Certification

Scheme (MCS) approval, if found guilty of intentionally misleading consumers. If the public

perception sways against PV installation, it could deliver a significant delay in the acceptance of

renewables as a potentially major energy source.

Muhammad-Sukki, et al. (2012) outlined a UK assessment for solar PV potential profits and future

installation rates, immediately after the April 2012 Feed-in tariff (FiT) changes. This study included a

financial analysis for a domestic property consuming 3,300 kWh per annum with a 2.6 kW PV

installation. The conclusion drawn was that return on investment (ROI) would fall to 4.2% from 10%

prior to the changes; this loss of profits would also cause an increase in the payback period (PBP) to

13.2 years. Other concerns raised included the potential loss of 4,000 to 25,000 jobs in the UK

through the slowdown of the PV market. This was all with the false assumption that the FiT rate

would only fall to 21p/kWh, as was promised in early government proposals. Instead, the FiT rate

fell to ≈ 12p/kWh), increasing the expected profit and job loss.

In February 2016, the FiT rate fell by a similar proportion to ≈ 4.4p/kWh for small scale PV

installations (Gauke, 2015). This project will investigate the potential effect of these changes on the

financial aspects for systems installed across the UK. At the time of research, online solar PV

assessment calculators still employed a FiT rate assumption of 12.4p/kWh for financial calculations.


2

The proposed roll out of SMART meters across the UK by 2020 will change the tariff payment

methods used to calculate export and FiT payments. Instead of a flat 50% export versus self-

consumption ratio, the true export value will be quantifiable and used when calculating export

payments; potentially reducing the lifetime profits of a PV system. A way to accurately predict

export payments is not possible without a full analysis of the assessed properties historical electricity

usage. Data supplied by EG-Audit (EG) will allow the discovery of usage trends for a range of

establishments. This data has been collected over several years from six case study properties

across the North West.

The development of a tool capable of comparing net electricity usage against predicted solar

generation will allow EG to offer bespoke quotes for their customers. Moreover, the tool will

contain the most recent government defined FiT rates and export tariffs. From the analysis of

several leading solar assessment calculators, EG request the developed tool be at least as accurate

as what is currently commercially available. In the interest of transparency, this report outlines the

calculation methods used in EG’s final assessment tool.

This project requires the review of applicable literature and research, investigation into the varying

performance of currently available assessment calculators, and comparison of these calculators

against the final tool developed for EG. Detailed knowledge of solar panel characteristics and the

effect of the varying operating conditions are important for accurate predictions. This research

hopes to highlight any novel variables and assumptions not employed by the current crop of

assessment calculators, for use in EG’s final tool. From the data supplied by EG, six separate solar

assessments were completed for case studies of differing property type.


3

1.2 Objectives

Research solar irradiance variables, understanding the scientific principles, and identify

methods for application in EG’s assessment tool.

Develop system to automatically interpret the half hourly usage data supplied by EG, then

compare and analyse the identified usage trends for each case study.

Assess and compare the required inputs for, and results presented by, the current crop of

commercially available solar assessment calculators.

Identify areas where the alternative calculators are insufficient, outdated, or inaccurate;

proceed to include recognised novel features in EG’s assessment tool.

Compare the results from the alternative calculators to find an error margin, within which

the final assessment tool should aim to lie.

Develop system that can predict PV generation levels from system specification information

with a half hourly format output.

Develop solar assessment tool that identifies export versus self-consumption ratios for the

use of EG auditors and their customers.

Discuss key findings from the developed assessment tool for each case study property type.

Justify the accuracy of the final assessment software by comparing and contrasting

predictions against error margins identified from the alternative calculator sample.

Identify areas that require further investigation and outline possible research methods for

future researchers.


4

1.3 Introduction

1.3.1 EG-Audit

EG is an energy auditing company based in Wervin, Cheshire. They offer their services to a wide

variety of customers ranging from a multi-national industrial supplies company, to private landlords

with several properties. Usually, EG deal with energy account management (for example by

detecting billing errors and notifying suppliers when mistakes have been made). After garnering a

reputation as an approachable and effective appraisal asset, EG aim to expand into other markets.

Moving into the future, EG anticipate specialisation in the energy project management market. This

plan includes the creation of a solar assessment tool for customers looking into the potential of PV

installation. It is the desire of EG for their tool to be at least as accurate as currently available solar

assessment systems, whilst remaining an entirely impartial third party to solar panel installations.

Past experiences have shown that solar quotations delivered to EG’s customers can sometimes be

well off the mark. With some solar suppliers quoting systems capable of delivering much more than

required, while others can be misleading with the use of subversive result presentations. In October

2014, EG were party to a quote that predicted a profit of £918,000 after 20 years for a 260kWp PV

installation. This was for a golf course simply wishing to cover their energy usage, encouraged to

invest £400,000. It is this form of questionable selling tactics that have led EG to feel the need for

such a study. It is their hope that this tool will factor in many variables that alternative assessment

calculators do not; allowing EG’s customers to receive a truer assessment of the potential of PV

installation for their properties.

1.3.2 Microgeneration Certification Scheme

“The Microgeneration Certification Scheme is an industry led certification scheme for microgeneration products and installation services. Supported by the Department for Energy and Climate Change (DECC), MCS seeks to build consumer confidence and support the development of robust industry standards. It provides confidence in the marketplace and wholly supports government policy within the microgeneration sector.” (Electrical Contractors Association, 2012).

This project will abide by the rules and regulations outlined by the MCS in the 2012 published ‘Guide

to the Installation of Photovoltaic Systems’ wherever possible. All solar panel companies must be

MCS approved in order to qualify subsequent sales and installations for export and FiT payments.

Accordingly, EG’s final tool should be suitable for MCS photovoltaic standard MIS 3002 approval.


5

1.3.3 Solar Power

Solar power encompasses any capture of energy from sunlight irradiated and reflected onto the

earth’s surface for conversion into heat or electricity. This can be done directly, by the application of

the PV effect, or in-directly, by using concentrated solar power systems (CSP). These forms of

energy generation are classified as renewable as they do not require the depletion of a fuel source

during conversion to heat or electricity.

A move away from the combustion of CO2 releasing fuel sources has become a matter of urgency;

particularly with increasing government carbon taxes and stricter intergovernmental agreements

pressuring the UK’s largest offenders. Solar power is seen by many as having the largest potential to

replace the tradition oil, coal, and gas fuelled power systems; with many projections claiming that

investments by BRICS nations will make solar “the world’s largest source of electricity by 2050”

(International Energy Agency, 2014).

1.3.4 Solar Thermal Energy

Solar thermal energy (STE) is the generation of usable energy by the transformation of the sun’s rays

incident on the Earth’s surface. STE can be further subcategorised into passive or active systems.

Small scale passive systems (such as Trombe walls, conservatories, and atrias) utilise natural light

and heat energy, without the need for mechanical/electrical devices or to compromise the intended

function of a building. They are simple and cheap in design, and require relatively little

maintenance. However, they can be expensive and/or difficult to retrofit onto older properties.

Small scale active systems (such as evacuated tube, flat plate, and unglazed solar water heaters)

utilise heat energy from the incident rays to serve water heating requirements. They typically have

very small operational and maintenance costs, however transfer fluids used in higher efficiency

systems can be toxic if leakage occurs.

These solar energy systems partially displace the use of electricity or gas for space/water

temperature control purposes. As of 2013, there was an installed capacity of 326 GW (REN21, 2014)

for STE systems similar to the one seen in Figure 1.3-1. Similar to CSP, these systems are also

dependent upon high levels of solar insolation to induce practicable space/water warming levels.

Some simpler water tank warming systems do not require electricity to function. As such, China

(with large swathes of rural areas without electrical grid connectivity) makes up 64% of the world’s

solar heating and cooling capacity (REN21, 2014), with the remainder consisting almost entirely of

equator lying countries.


6

Figure 1.3-1: Evacuated tube solar heating and cooling system (Equilibre Personnel, 2016).

CSP systems make use of lenses and mirrors to direct sunlight from a large area onto a high

efficiency conversion point or points. The concentrated energy is captured as thermal energy that

warms a ‘working fluid’, typically molten salts (Hasuike, Yoshizawa, Suzuki, & Tamaura, 2006). The

fluids are heated to 150°C -350°C, before flowing into a boiler to drive a steam turbine; typical of

other power generation plants. Parabolic trough systems (such as the one seen in Figure 1.3-2) are

the most frequently used CSP system around the world; with > 3,400MW installed worldwide as of

2013.

Figure 1.3-2: Parabolic trough concentrated solar power plant near in California (SkyFuel, 2011).

Fresnel linear plate reflectors work in a similar fashion, capable of reaching higher efficiencies due to

the higher reflectivity of the mirroring system (Zhu, Wendelin, J, & Kutscher, 2013). Alternatives

such as Stirling engines and solar power towers are also well developed (REN21, 2014). A large scale

example of a point focus tower is the 20MW PS20 consisting of 1,255 heliostats focused towards a

531 foot high tower (NREL, 2015).


7

The advantages of CSP systems include their higher conversion efficiency when compared against

the current crop of commercially available PV systems, as well as higher storage capabilities than

their PV counterparts. However, these systems are extremely dependent upon high levels of solar

insolation to be reasonably practicable. Hence, locations such as Spain and southern USA make up

over 70% of the world’s installed CSP capacity.

1.3.5 Photovoltaic Solar Energy

PV is the form of solar power of concern in this project. PV converts sunlight directly into electricity

via the application of the photovoltaic effect. Recent figures estimate an installed capacity of

178GW for PV powered systems worldwide (EPIA, 2014). As of the end of November 2015, overall

UK solar PV capacity stood at 8,437MW derived from 815,505 installations (Department for Energy

& Climate Change, 2015). DECC also state is this report that “capacity eligible for Feed in Tariffs

stood at 3,615MW across 796,884 installations. This represents 43 per cent of total solar capacity

and 98 per cent of all UK PV installations”. The UK lies fifth in the list of annual investment in PV

technologies worldwide published by EPIA; with only China, Japan, the USA, and Germany ahead

respectively. When this list is adjusted for installed capacity per capita, however, Germany is the

world leader by a significant margin. In regards to the EU, Germany accounts for over half of all

installed PV capacity (as seen in Figure 1.3-3).

Figure 1.3-3: 2010 projected to 2015 EU PV market share (adapted from Vatansever, et al., 2012).

Unlike other forms of solar power, PV does not require high levels of solar insolation to convert a

practicable level of energy into usable electricity. For this reason, the benefits of installation can be

seen almost worldwide (Perez, Ineichen, & Seals, 1990). The key drawback for this technology is

that production can only take place during daylight hours; therefore a form of energy storage or

UK, 0.3% Spain, 3.0%

Rest of EU, 3.3%

Italy, 18.0%

Germany, 56.0%

France, 5.0%

Czech Republic,

11.0%

Belgium, 3.0%


8

transportation is essential if round the clock supply is to be attained. PV systems range from small

scale roof mounted installations to vast field arrays capable of utility scale output in the gigawatts

(see Figure 1.3-4). Small scale, grid connected systems account for a large majority of the UK PV

market. By feeding unused electricity back into the grid, these systems ensure that very little of the

produced energy is wasted.

Figure 1.3-4: Example PV systems from small to utility scale (from left to right: roof top ≈ 4 kW, field array ≈ 1-5 MW, solar power station < 1 GW)

France, as of February 2016, announced plans to trial a 1000km stretch of ‘solar roadway’, expected

to deliver 8% of France’s electricity demands by 2020 (Guerrini, 2016). Solar roadways apply PV

technologies either in place of or on top of the road surface. The treatment required to make the

cells weather/traffic proof leads to a substantially higher purchase cost than typical PV panels

(estimated £4/kWp for the announced Wattway project). This is before considering the

astronomical supplementary equipment and operational costs associated with a new underground

grid system. Moreover, the efficiency of these panels under standard test conditions peaks at 15%;

compared to 18-19% for conventional PV systems. This is assuming ideal conditions, before

factoring in dust and mud losses accumulated from the passing road vehicles. South Korea has,

perhaps, found a more sensible solution by installing PV modules to act as shading for bike lanes that

run along the sides of roads (see Figure 1.3-6).

Figure 1.3-5: Colas Route solaire Wattway (Guerrini, 2016).


9

Figure 1.3-6: South Korea’s solar bike lane shading system (Alter, 2015).

Due to recent exponential growth in the PV installation market, costs are falling dramatically.

Swanson’s law is a proposed relationship predicting PV costs over time. It predicts the cost per watt

falls by 20% for every doubling of the cumulative shipped volume of PV modules. So far, this

observation fits the true cost of PV modules very well, as seen in Figure 1.3-7.

Figure 1.3-7: Swanson’s Law of experience analysis applied to PV module cost per watt (adapted from SEMI, 2014).


10

At present installation rates, the cost of PV systems halves every 10 years or so. For silicon PV cells,

the most common type of module used, prices have fallen to ≈ 24p/W, from £50 per Watt in 1977

(Swanson, 2006). The International Energy Agency (2014) state that the actual module now

accounts for less than 40% of the total installation cost; ancillary components and soft costs now

make up the majority of any quote (e.g. acquisition, labour, and financing costs). When everything is

taken into account, 2014 costs for UK installation were given as follows (all prices taken from

International Energy Agency, 2014);

Residential Property- £1.85 per Watt

Commercial Property- £1.60 per Watt

Utility Scale- £1.20 per Watt

The dramatic fall in PV cost is one of the key reasons why 98% of PV capacity has been installed since

2004; moreover 85% since 2011. In spite of this, there has been a recent decline in European

markets, barring a few nations such as the UK. With government budget cutbacks towards

renewable energy since the economic crash of 2008, there had been an unprecedented rush to

install before the expiry of the generous Feed-in-tariff (FiT) offered by the UK government. 2014

saw an 80% increase in cumulative PV capacity from the year before, with a similar increase

expected over 2015 (Migo, 2014). This led to a total PV capacity of 8.6 GW across the UK by 2016

(value taken from Figure 1.3-8).

Figure 1.3-8: Projected global cumulative capacity up to 2019 (adapted from EPIA, 2014).

0

100

200

300

400

500

600

2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

Glo

bal

Inst

alle

d C

apac

ity

(GW

)

Cumulative Capacity

Optimistic Scenario

Conservative Scenario


11

Figure 1.3-9: Global grid parity breakdown as of 2014 (adapted from Shah, et al., 2014).

Possibly the largest obstacle in the adoption of PV systems on a large scale is the lack of local grid

parity. Grid parity is defined as when an energy source “can produce power at a levelised cost of

electricity (LCOE) to the conventional electricity grid”. Very few places have reached this point for

solar PV (seen in Figure 1.3-9). LCOE is calculated by dividing the capital and maintenance costs of

the panels over their anticipated lifetime by the estimated production over the same time period

(e.g. a £7,000 4kW system produces 25,000kWh of electricity over a 20 year period ∴ LCOE ≈

28p/kWh). This oversimplified calculation also assumes that the panel purchaser will remain the

beneficiary of the energy production for the full lifetime of the panels. Whereas, the average length

of time that occupants in England have lived at their current address is 8 years and falling (Randall,

2011).

The most commonly used material for PV systems is silicon. Silicon is a group IV element and forms

the basis for the vast majority of semiconductor materials currently used in industry. The long

history of silicon use for integrated circuitry, and its historical application to solar equipment, makes

it the most popular PV technology currently commercially available.


12

1.3.6 PV system equipment

The key components required for PV systems are as follows;

Solar Panels Inverter Generation Meter

Light collected by the panels

generates a direct current (DC)

Converts DC into alternating

current (AC) for application

Records all electrical units

produced by system

Fuse Board Utility Meter Monitoring System

Electricity is distributed to

appliances in the home where

required

Logs the electricity taken and

exported to the grid when bi-

directionally equipped

Most modern systems allow

real time generation and

consumption monitoring

Many other ancillary components, such as brackets and wiring, are also required; however their cost

is very little in comparison to the components listed above. Efficiency losses are seen within each of

the listed components; therefore regular maintenance is required to ensure the system is running to

a practicable efficiency throughout module lifetimes. Inverters cause the largest losses of all the

above listed components, with modern commercially available inverters still only operating at 92-

96% efficiency (Vignola, Mavromatakis, & Krumsick, 2011). Vignola, et al. also state that peak

inverter efficiency is seen for solar irradiances of roughly 600 W/m2, this is typical for non-summer

month daylight in the UK.

1.3.7 Alternative PV options

Despite its market dominance (as seen in Figure 1.3-10), silicon is not the only material available for

use in PV solar cells. Several other semi-conductors, dye sensitised metal oxides, polymers, and

organic compounds are capable of achieving efficiencies similar to, or in excess of, silicon. NREL

publish an annual chart outlining efficiencies achieved in laboratory conditions by facilities around

the world. It uses recently published papers to assess the progress of solar technologies, classifying

them into five key areas (Multi-junction cells, Single-junction GaAs, Crystalline Si cells, Thin-film

technologies, and emerging technologies). Figure 1.3-11 is the most recent issue of the cell

efficiency chart, with record efficiencies of 44.7% held by four-junction or more (concentrator) solar


13

cells. For many technologies it can be seen that their efficiencies have remained relatively stagnant

since the late 1990s. This may be because the physical limitation of the solar cell has been reached,

or because of a drop off or complete removal of funding for this material. Historically, the private

funding of universities or dedicated laboratories tend to make the largest advancements in the

technology, rather than government investment.

The potential of alternative materials (particularly perovskite solar cells) in replacing silicon PV is

increasing, due to their lower relative cost and availability; recent advances in experimental

efficiencies are starting to match those of conventional silicon systems. One key advantage for

materials such as Perovskite is the simple processing and manufacture of the functioning PV module.

Spray coating has been touted as a possible manufacture method for perovskite cells; research is

currently investigating the scope of this highly scalable, cheap, and efficient process (Barrows,

Pearson, Kwak, Dunbar, Buckley, & Lidzey, 2014).

Figure 1.3-10: Global market share by PV technology from 1990 to 2013 (adapted from Fraunhofer Institute for Solar Energy Systems, 2015).

In recent years, Perovskite solar cells have shown great promise. Starting at an unremarkable

efficiency of only 3.8% during exploratory testing in 2009, it has raced to a competitive 20.1%

efficiency as of 2015 (Kojima, Teshima, Shirai, & Miyasaka, 2009). Advances of this speed have not

been witnessed since two-junction (concentrator) cells in the early 1990s. Whereas this level of

advancement could not be sustained in the past, Perovskite appears to be a major hope in reducing

high efficiency solar cell costs (Collavini, Völker, & Delgado, 2015). The key benefit of this material

coming in its relative abundance, hence low raw material costs. However, the key obstacle to

overcome for Perovskite lies in the removal of poisonous lead from the structure of the compound

without affecting the promising efficiencies seen to date (Noel, et al., 2014).


14

Figure 1.3-11: NREL multi-strand best research cell efficiencies chart (NREL, 2015).


15

1.3.8 How does Silicon PV work?

Light is made of packets of energy call photons. The energy contained in photons is what drives the

photovoltaic effect. The energy of any photon can be determined by;

𝐸 = ℎ𝑣

Equation 1.3-1

Where E = energy, h = Planck’s constant (6.63 × 10-34 Js), v = frequency (calculated by 𝑐 = 𝑣𝜆), c =

speed of light (3 × 108 m/s), and λ = wavelength.

Figure 1.3-12: General arrangement of silicon PV cell.

Put simply, the operation of any PV cell requires the absorption of light (whether it be sunlight or

artificial light) creating electron-hole pairs, followed by the separation of the generated charge

carriers, and the extraction and passing of the charge through an external circuit. This is done by

layering two or more sheets of semi-conductive material that have been doped in order to induce a

positive or negative electric charge in each respective layer. Typically, phosphorous is used to create

negative charge by adding electrons to the layer (n-region), whilst boron is added to create the

positive charge (p-region); this creates an electric field at all the points where the two layers touch

(cell junction). Electron-hole pairs are created along the junction when electrons from the n-region

move to fill the positive space between the two layers. When a photon hits electrons in the silicon

junction, the absorbed energy will excite them from their usual state, returning electrons to the n-

region, creating a charge imbalance. The electrons will attempt to return to their electron-hole

pairs; however the synthesised electrical field does not allow them to do so. Instead, by connecting

the two layers, forming a closed circuit, a direct current can be induced over the apparatus. The


16

flow of electrons is constant when photons of high enough energy are directed towards the cell

surface. This general explanation can be seen graphically in Figure 1.3-12.

An antireflective coating is used to reduce the amount of sunlight reflected away from the silicon

surface, potentially improving overall efficiency by 3-5% (Alves & Boling, 2010). By optimising the

refractive index of the coating, up to 30% less sunlight is reflected when compared to bare silicon

(Green M. A., 1995).

1.3.9 Bandgap

The bandgap is the minimum energy needed for electrons to jump to higher valence energy bands

from their bound state. Only semi-metal, semi-conductor, or insulator materials possess valence

energy levels that display band gaps. The valance energy levels for metals overlap, meaning

electrons can freely move between valence bands without the opportunity for charge separation.

When the surface of a material is hit by an incident photon, the electrons can participate in

conduction by excitement across the bandgap. The amount of energy generated by a cell is

dependent upon the bandgap as it dictates the amount of energy required by impacting photons.

This leaves behind a positive ‘hole’ in the semi-conductor material. Electrons will flow freely in the

material attempting to fill this ‘hole’ (the electron and the ‘hole’ are deemed the charge carriers).

The conduction band, EC, is the energy level at which electrons are in their free state, whereas Ev is

known at the valence band, where the difference between the two is the bandgap EG. Different

cases for incident photon energy, EPhoton can be seen in Table 1.3-1;

Table 1.3-1: Situational analysis for different photon energy levels.

EPhoton < EG

Photon passes through semi-

conductor, as if transparent No energy generated

EPhoton = EG Efficient absorption, with no heat

wasted Efficient energy generated

EPhoton > EG Electron-hole pair generated, but

electron is thermalised

Energy generated, thermalised

energy wasted

The absorption coefficient is equal to the distance into a material light can travel before being fully

absorbed. This is a factor of the material itself and the wavelength of light. The absorption

coefficient of silicon can be seen in Figure 1.3-13. Silicon has a relatively high absorption coefficient

(Green M. A., 1995), making the excitement of electrons into the conduction band easier. Blue light,

with a much smaller wavelength than the rest of the optical range is absorbed over a relatively short

distance, whereas red light, at the other end of the spectrum is absorbed over much larger

distances. However, for silicon this is still a matter of a few hundred microns.


17

Figure 1.3-13: Absorption coefficient of silicon against wavelength on a logarithmic scale (data collected by and adapted from Green, 1995).

Following absorption, the electron-hole pair is meta-stable and can only exist for a period of time

known as the ‘minority carrier lifetime’ (a matter of femtoseconds for metals, but nanoseconds for

materials that possess a bandgap). A second process must occur to prevent recombination, spatially

separating the electron-hole pair at the electric field of the p-n junction. By connecting the emitter

and base of the solar cell, the carriers will travel through the external circuit, inducing a current.

Figure 1.3-14: Bandgap and associated maximum efficiency for single layer silicon p-n junction as outlined by and adapted from Shockley and Queisser, 1961.

The optical bandgap determines the portion of photons in the solar spectrum a photovoltaic cell can

effectively be converting into electricity. The Shockley-Queisser limit outlines the maximum

theoretical energy a typical p-n junction solar cell can produce. As silicon has a bandgap of

approximately 1.1eV between 250K – 350K (Kittel, 1986), it results in a maximum potential efficiency

of 32% for a single layer Silicon structure (Shockley & Queisser, Detailed Balance Limit of Efficiency

1.0E-08

1.0E-06

1.0E-04

1.0E-02

1.0E+00

1.0E+02

1.0E+04

1.0E+06

200 400 600 800 1000 1200 1400

Ab

sorp

tio

n c

oef

fici

en

t (c

m-1

)

Wavelength (nm)


18

of p-n Junction Solar Cells, 1961). When a theoretical multiple layer solar cell is considered with an

infinite number of layers, and concentrated sunlight is applied, the maximum efficiency approaches

86% (Vos, 1980).

1.3.10 Recombination

The lifetime of silicon cells is largely dependent upon the recombination rate of the electron-hole

pairs (Shockley & Read, Statistics of the Recombinations of Holes and Electrons, 1952).

Recombination is when the electron returns of the valence band without inducing a current across

the p and n junctions. There are several forms of recombination in solar cells, leading to losses in

efficiencies.

Radiative recombination is when an excited electron returns to the valence gap releasing a photon ≈

EG, conversely this is negligible in silicon as it is deemed an indirect bandgap semiconductor (Werner,

Kolodinski, & Queisser, 1994). Shockley-Read-Hall (SRH) recombination occurs when impurities are

present in the semiconductor material. The impurities give rise to an extra energy level between the

valence and conductor gaps. Excited electrons will first fall to this extra level, then eventually to the

original valence band; this releases two smaller photons, the sum of which is equivalent to EG.

Whilst this form of recombination becomes increasingly significant with the increasing age of the

solar cell, Auger recombination is the most predominant form of recombination in silicon cells.

Auger recombination ultimately dictates the lifetime of the cell; it is also most prevalent in heavily

doped silicon cells. When excited electrons return to the valence band, instead of releasing a

photon, the energy is transferred to another electron in the conductor band. This electron will

slowly thermalise as it returns to its own original energy level (Auger, 1923).


19

1.4 Electricity Metering

1.4.1 Metering Point Administration Number

The metering point administration number (MPAN) is a unique 21 digit number used by electricity

suppliers to identify individual properties. Figure 1.4-1 outlines the meaning for all digits for any

MPAN.

Figure 1.4-1: Typical printed format for MPAN number included on most electricity bills.

Profile classes define how electricity is expected to be consumed throughout the day by a property.

Split into domestic or non-domestic profiles, all the case study properties supplied by EG would be

classed 00 or 03 and higher.

Table 1.4-1: Profile class definitions.

Profile Class Definition

00 Half-hourly supply

01 Domestic Unrestricted

02 Domestic Economy 7

03 Non-Domestic Unrestricted

04 Non-Domestic Economy 7

05 Non-Domestic Maximum Demand 0-20% Load Factor*

06 Non-Domestic Maximum Demand 20-30% Load Factor

07 Non-Domestic Maximum Demand 30-40% Load Factor

08 Non-Domestic Maximum Demand >40% Load Factor

*Load factor is the ratio between unit consumption and maximum demand.

Electricity unit prices vary throughout the day, week, and year. Units are classified into the

categories seen in Table 1.4-2. For single unit rates (such as unrestricted plans) there need only be

one register. Registers are defined as points where electrical units switch in cost from one point in

time to another. For pricing structures other than unrestricted, the meter will need to switch over

to the new unit pricing during that period of time. Seasonal Time of Day (STOD) has many different


20

prices for units throughout the year as well as time of day. Typically these structures are reserved

for large electricity consumers, with a profile class of 06 or higher.

Table 1.4-2: Unit categories and their average UK defined periods of use.

Pricing Structure Unit Time Period

Unrestricted All Anytime of day and year

Day/Night Day 7AM-12AM every day of the week

Night 12AM-7AM every day of the week

E/W Weekdays 7AM-7PM weekdays

Evening/Weekends 7PM-12AM weekdays, all day weekends

W/N/E/N

Weekdays 7AM-7PM weekdays

Evening/Weekends 7PM-12AM weekdays, 7AM-12AM weekends

Night 12AM-7AM every day of the week

The Meter Time Switch Code (MTC) indicates how many registers (sets of meters reads or dials) a

property’s electricity meter has (Energy Linx, 2016). This depends on pricing structure agreed with

the distribution network operator (DNO). Table 1.4-3 displays the vast number of MTC options

available for suppliers to define a property. EG possess a document that lists all MTC values and the

pricing structures they represent. In its current state EG’s assessment tool uses the average unit

price for each DNO from electricity bill calculations supplied by the DECC. However, a potential

future addition to the tool could be to include the, EG controlled, MTC document in order to

calculate annual electricity bills using half-hourly data and MPAN number only. This would save the

user from finding these values themselves before using the tool, creating a much more ‘user

friendly’ experience.

Table 1.4-3: MTC values and their definitions (Energy Linx, 2016).

MTC Definition

001-399 DNO Specific

400-499 Reserved

500-509 Codes for related Metering Systems- common across the industry

510-799 Codes for related Metering Systems- DNO specific

800-999 Codes common across the industry

The distributor ID defines the distribution area boundaries the property for assessment falls within

(see Figure 1.7-1). The MTC, profile class, and distributor ID are the three values of interest in this

project. These values define the rate at which electricity is charged throughout the day and year,

allowing the estimation of potential electricity before and after PV installation. The total number of

combinations for these three values exceeds 100,000. For ease sake, only the four main pricing

structures seen in Table 1.4-2 will be used for tariff change suggestions, with all STOD rates


21

excluded; this still left 60 separate pricing plans to be including in the assessment tool. The unit

prices for the different pricing structures have been averaged from 2015 distributor contracts

supplied by EG (seen in Table 1.4-4). Properties with automatic meter readings will be charged the

quarterly AMR rate also seen in Table 1.4-4, all other properties will be charged the Non-AMR rate.

For all case studies supplied by EG, the Non-AMR rate was used for tariff calculations.

Table 1.4-4: Pricing structure unit prices used for tariff suggestion aspect of tool (EG-Audit, 2016).

Pricing Structure

Billin

g Perio

d*

Automatic Meter Reading Standing

Charge (£) Units (p/kWh)

AMR Non AMR A

ll

Day

Nigh

t

Weekd

ays

Evenin

g

Weeken

d

Unrestricted Q 23.3 23.3 9.96 - - - - -

Day/Night Q 23.3 23.3 - 10.4 6.15 - - -

E/W Q 35.2 16.6 - - - 10.0 8.5 8.5

W/N/E/W Q 35.1 16.5 - - 6.17 10.8 9.7 9.7

*Billing period can be monthly or quarterly, in this case all tariffs were quarterly

The Line Loss Factor (LLF) code estimates the expected costs charged to a property’s electricity

supplier, due to energy lost in transit from the supply network to the meter. The meter point ID

number is simply a unique number within the distribution area used to identify individual property

meters. The check digit is calculated from the distributor ID and meter point ID number for external

system use. However, these values are irrelevant to electricity rates and are of no concern for this

report.

1.4.2 Half Hourly Data

The data for analysis is supplied by EG in half hourly format (see Figure 6.3-1). kWh/half hour

readings are taken every 30 minutes and recorded by specially installed meters. The readings are

numbered chronologically throughout the day (e.g. 1 = 00:00, 2 = 00:30…, 48 = 23:30). The

properties selected for analysis have each provided at least one year’s worth of continuous data for

trend identification; without this amount of data, the tool would not be able to successfully model

yearly trends in electricity usage. This equates to a minimum of 17,250 continuous data points per

property. Over 70,000 data points have been provided for certain case studies, reducing overall

potential for error during trend setting and usage predictions.


22

1.4.3 Property Types

EG has kindly provided over 15 years worth of electricity usage data for several property types. For

data protection purposes, these properties have been renamed as the property classification they

were deemed to match most accordingly (see Table 1.4-5). Each property type has been classified

according to their operating hours and occupant intentions.

Table 1.4-5: Property classifications agreed with EG for analysis in the assessment tool.

Classification Examples Qualifying Criteria

Commercial Office buildings, Retail properties Typical 9-5 working hours, closed on

weekends

Educational

Primary schools, Secondary

schools, Colleges, University

buildings

Typical UK school opening hours, closed on

weekends, large periods of shutdown

throughout year

Industrial Factories, Manufacturing centres 24 hour operation, continuous energy

intensive processes throughout the year

Light

Commercial

Offices and Retail properties with

flexible working hours Atypical working hours, closed on weekends

Residential Block of flats, Hotel, Apartment

Complex Property must be occupied during test period

Social Golf Clubhouses, Pubs, Bars,

Nightclubs

Late night opening hours, open all days of the

week

EG provided the following background information on the properties;

Commercial- Small office building, typical 9-5 working week.

Educational- Small primary school with several, poorly insulated, mobile buildings.

Industrial- 24 hour industrial operation and business office with typical 9-5 working day and

Monday to Friday working week.

Light Commercial- Family run business with atypical working hours.

Residential- Student accommodation apartment complex.

Social- Golf course club house, open to members all year round.

There are several properties that do not fall within the constraints described in Table 1.4-5.

‘Assembly and Leisure’ is a key example encompassing properties such as museums, libraries,

gymnasiums, or religious institutions. Unfortunately, access to usage data for these properties was

not available during the research period. These properties were also deemed to use electricity in

such vastly different fashions that only bespoke assessments would most likely harbour accurate

results, rather than attempting to apply unsuitable usage trends.


23

The eventual goal of EG is to analyse enough historical electricity usage for different properties to

develop a ‘trend’ for different property types. At least one property of each classification from each

distribution zone should give a suitable sample size for an accurate representation of energy usage

patterns throughout the day, week, and year across the UK. Currently, EG has been able to provide

electricity usage data for one property of each classification from the North West distribution area.

A potential future study could include the collection and analysis of further data to see whether the

trends developed for use in the assessment tool were suitable for the whole of the UK.


24

1.5 Solar PV Generation Factors

1.5.1 Insolation/Irradiance

Solar insolation is the amount of electromagnetic radiation incident to the earth’s surface, typically

given in kWh/m2/day. Insolation levels vary throughout the day and year, depending upon the

number of daylight hours. Midday during peak summer months on a clear day is when solar

irradiance is at its highest. If the solar insolation is known for a particular area, the amount of

energy a solar collector will generate can be calculated (when system losses and further factors are

considered). Figure 1.5-1 establishes the yearly sum of irradiance in kWh/m2 across the UK from

averaged historical data. This is equivalent to the energy generated by a 1m2 panel if it is 100%

efficient. Because no system is inherently lossless, any installed system will be designed to a

kilowatt peak (kWp) value. kWp factors in system losses and acts as the anticipated generation for

the area of cells including in the quotation. This is a nominal value specified under standard test

conditions (STC); 1000 W/m2, 25˚C cell temperature, 1 m/s wind speed, air mass = 1.5 (Green M. A.,

2009).

Figure 1.5-1: Yearly total irradiation in kWh/m2 for UK (averaging period 1993-2007, information courtesy of the Met Office).

A general trend that can be seen in Figure 1.5-1 is that irradiance decreases as latitude increases.

Scotland, for example, receives 350kWh/m2/annum less solar energy than Southern England on


25

average. Coastal regions also receive far greater levels of radiation than areas further inland, due to

differing weather patterns. For this reason, accurate geo-positioning is required for the accurate

solar assessment of any case study. MCS specifies solar insolation values for 21 zones across the UK

that all solar assessment tools must use. Four zones are further split into two smaller areas, due to

significant variations in annual insolation levels (see Figure 1.5-2).

Figure 1.5-2: MCS postcode specified regions for which irradiance values have been averaged according to Met. Office data collected from 1993-2007 (Electrical Contractors Association, 2012).


26

Table 6.3-1 outlines the zone data that used to identify the correct MCS zone upon postcode entry of

the location for assessment. The tool then reads the expected generation for the specified location

depending upon the system inclination and orientation (see Figure 6.3-2).

The data supplied by the MCS is given in annual kWh/kWp. Data gathered from 1993-2007 by the

Met. Office is used in alternative calculators to find the total lifetime generation by multiplying by

expected panel lifetime. For alternative tools this value is split evenly between self-consumption

and export and the associated savings/revenues are applied to find the lifetime profits for an

installed system. This assumption will result in an overestimation of lifetime profits for properties

that export the majority of the electricity that they generate (as self-consumption savings per kWh

are in excess of export revenues). Even though profits are underestimated for the majority of

properties that export less than 50% of their total generation, the assessment cannot be deemed

accurate. This is the key advantage of EG’s tool over what is currently available.

1.5.2 Seasonal Variation

Figure 1.5-3: Monthly generation as a proportion of annual generation for the UK; adapted from historical data gathered by Electrical Contractors Association (2012).

Due to the limited number of daylight hours, combined with seasonal weather patterns, it is

apparent that winter days will not generate as much electricity when compared against summer

days. For the purposes of the assessment tool, the distribution seen in Figure 1.5-3 will be used for

yearly generation predictions. This graph highlights the importance of utilising generation during the

summer months wherever possible. For example, average generation during May is equivalent to

0%

2%

4%

6%

8%

10%

12%

14%

16%

Po

rpo

rtio

n o

f an

nu

al G

en

era

tio

n


27

more than seven December’s worth of generation. Over 50% of total annual generation is

generated during only four months of the year (May to August). This further stresses the

importance of seasonal usage analysis.

1.5.1 Sun Path

The apparent motion of the sun has a significant impact on the amount of insolation incident on the

Earth’s surface per day, through the year. For this reason, accurate modelling of sunrise/sunset

times, and number of daylight hours throughout the year is required for any self-

consumption/export ratio dependant solar assessment.

The equations used to model yearly sunrise and the sunset times are as follows. The number of

daylight hours is equal to the difference between the results of these two equations;

𝑆𝑢𝑛𝑟𝑖𝑠𝑒 = 12 −1

15˚𝑐𝑜𝑠−1(− tan 𝜑 tan 𝛿) −

𝑇𝐶

60

Equation 1.5-1

𝑆𝑢𝑛𝑠𝑒𝑡 = 12 +1

15˚𝑐𝑜𝑠−1(− tan 𝜑 tan 𝛿) −

𝑇𝐶

60

Equation 1.5-2

Where ϕ is the property latitude (˚), δ is the declination angle (˚), and TC is the time correction factor

(mins).

The declination angle varies seasonally due to the tilt of the Earth and can be found using;

𝛿 = 23.45˚ sin [360

365(𝑑 − 81)]

Equation 1.5-3

Where d is the day number (1 = 1st January, 365 = 31st December, etc.). 81 is subtracted from d as

this is the number of days between the 1st January and when the sun reaches a solar noon latitude of

0˚ (solar equinox) on the 22nd March. The Earth is tilted by 23.45°, and as such δ will reach this value

at both the summer (+23.45˚) and winter (-23.45˚) solstices. 360/365 represents the angle by which

the Earth travels round the sun each day; this is equivalent to 0.986° per day.

TC accounts for the variation of the local solar time within a given time zone. For countries like

China that operate on a single time zone, whilst longitude varies by as much as 60˚, TC becomes a

very important factor during assessment. Even for the UK the local solar time can vary by upwards

of 45 minutes from GMT. TC is found using the following equation;


28

𝑇𝐶 = 4(𝐿𝑜𝑛𝑔𝑖𝑡𝑢𝑑𝑒 − 𝐿𝑆𝑇𝑀) + 𝐸𝑜𝑇

Equation 1.5-4

Where longitude is given in degrees, LSTM is the local standard time meridian, and EoT is the

equation of time. LSTM is a reference meridian used for different time zones converting the

difference between local time and GMT to longitudinal degrees from Greenwich a given location is.

For all areas of the UK (where GMT is local time), LSTM is equal to zero. The factor of 4 is derived

from the fact that the Earth rotates ≈ 1˚ every 4 minutes.

The EoT (mins) is an empirical equation the corrects for the eccentricity of the Earth’s orbit and axial

tilt (PVCDROM, 2016);

𝐸𝑜𝑇 = 9.87 sin(2𝐵) − 7.53 cos(𝐵) − 1.5 sin(𝐵)

Equation 1.5-5

Where B (˚) is dependent upon d (number of days since the start of the year) according to the

following equation;

𝐵 =360

365(𝑑 − 81)

Equation 1.5-6

A plot of EoT versus d can be seen in Figure 1.5-4. It can be seen that EoT is greater than 15 minutes

at certain points in the year, therefore correcting for EoT is essential for accurate modelling.

Figure 1.5-4: EoT versus d values used to model the path of the sun for generation calculations.

-15

-10

-5

0

5

10

15

20

0 50 100 150 200 250 300 350

EoT

(min

s)

d


29

An example application of all the above equations can be seen in Figure 1.5-5. The summer and

winter solstices can be seen at d = 172 (21st June) and d = 355 (21st December) respectively. The

location selected (CH2 4DW) fell within the boundaries zone 7E. The centre point for zone 7E has

positional latitude of +53.5˚ and longitude of +2.6˚. From the MCS post code zone data, the

longitudes and latitudes of the property can be found using Table 6.3-2.

Figure 1.5-5: Seasonal variance of sunrise and sunset times for CH2 4DW (+53.5˚, +2.6˚).

The relationship between time of day and proportion of daily generation is dependent upon the

relative position of the solar noon. As mentioned previously, the day has been split into 48 half

hourly periods. The time difference between each half hour of interest to the solar noon of each

individual day of the year will be squared to distribute the total daily generation throughout the half

hourly periods. An example of this distribution can be seen in Figure 2.2-1. Without this quadratic

relationship there would be no variation in irradiance levels as the sun approaches and retreats from

apparent solar noon. This would lead to a flat line of generation and inaccurately mimic the true

daily irradiance levels. This variance is vital when comparing against half hourly usage levels;

without it the export ratio estimations would be no more correct than what is currently applied.

1.5.2 Pollution Levels

Key variables in PV efficiency include the wavelength spectrum and intensity of the photons incident

to the cell surface. These are also values determined by the air mass (AM) of a particular location.

In order to compare different PV technologies, testing is completed at an AM of 1.5 and cell

temperature of 25°C. AM is defined as the ratio of a path length of light through the earth’s

atmosphere to a particular location relative to the length if the light had travelled vertically through

0

4

8

12

16

20

24

0 50 100 150 200 250 300 350

Ho

ur

of

Day

d

Sunset

Sunrise


30

the atmosphere. The equation to calculate AM at a given distance from the equator when

accounting for the curvature of the earth is as follows;

𝐴𝑀 =1

𝑐𝑜𝑠 𝜑 + 0.50572(96.07995 − 𝜑)−1.6364

Equation 1.5-7 (Kasten & Young, 1989)

Where ϕ is the latitude in degrees for the assessed location, the range of latitude of interest to the

assessment tool is 50° - 60° N.

Gases and particulates present in the atmosphere attenuate the light by scattering, absorption, or

reflection. O3 absorbs some UV in the stratosphere; H2O, CO2, NOX, and even O2 all contribute to the

attenuation of different wavelengths of light, as seen in Figure 1.5-6. H20 and CO2 are the key

contributors to the attenuation of higher wavelength energies of photons. Silicon cells are relatively

unaffected by these processes as the radiation absorbed tends to be of high energy and does not

match the bandgap of Si. The removal of these superfluous energies increases the overall efficiency

of solar cells when considering the output power relative to the incoming solar energy ratio.

Figure 1.5-6: Solar radiation spectrum for AM1.0, highlighting absorption bands for various atmospheric gases (adapted from Rhode, 2016).


31

Using the equation laid out by Kasten & Young in 1989, the UK ranges from AM = 1.6 at 50° latitude,

to AM = 2.0 at 60° latitude, however, this is before factoring in large atmospheric efficiency

inhibitors such as cloud coverage, photochemical smog, and ground level temperature variance.

Solar intensity decreases with increasing AM values in a non-linear fashion. Almost all high energy

photons are removed from sunlight between AM0 and AM1.0, nonetheless the remaining light

energy attenuation level, from equator to pole, can vary by as much as 70%. An attempt to

approximate light intensity for differing AM is laid out by Meinel & Meinel in 1976;

𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦, 𝐼 = 1.1 × 𝐼𝑂 × 0.7(𝐴𝑀)0.678

Equation 1.5-8 (Meinel & Meinel, 1976)

Where 𝐼𝑂 is the intensity of light reaching the earth’s atmosphere (1353 kW/m2) and the diffuse light

impacting the panel is roughly an additional 10% of the direct component. For polluted air (as is

seen in cities with large population densities) the exponent to which AM is risen is 0.715, whereas

for clean air it is 0.618 (American Society for Testing and Materials, 2012).

The AM values change light intensity levels for the latitudes of concern by as much as ±27%. EG’s

tool classifies locations into three types. ‘Large City’ is deemed the most polluted (AM raised to

0.715), ‘Village/Town’ the next most polluted (AM raised to 0.678), and ‘Countryside’ the least

polluted (AM raised to 0.618). These exponents were applied to the latitudes, and resultant AM,

values of interest to the UK. It was found that on average, living in a ‘Large City’ caused a 5%

reduction in overall irradiance relative to ‘Countryside’ levels; ‘Village/Town’ living induced a 2.5%

drop in irradiance.

1.5.1 Panel Inclination & Orientation

Panel Inclination is deemed the angle of the module surface from the horizontal plane. For the vast

majority of solar installations, this is dependent upon the pitch of the roof of the property in

question. However, for free standing field arrays the inclination can be optimised to receive the

maximum level of insolation throughout the day and year. 80% of properties in the UK fall within a

roof pitch range of 39˚ - 49˚ (Fewins, 2012). However, the insolation data provided by the Electrical

Contractors Association (2012) limits the acceptable inclination for modelling between 0˚-45˚.

Without further information the assessment tool will also be constrained by these limitations. A

future study may look into identifying a relationship between panel inclination and insolation levels

to allow the modelling of higher roof pitch properties.


32

Figure 1.5-7: Optimum inclination angle across the UK (adapted from Midsummer Energy, 2016).

Orientation is the angle of the module surface relative to due South. ±180˚ is due North, +90˚ is due

West, and -90˚ is due East. For rooftop installations only South, South West, or South East facing

roofs are considered suitable, unless the inclination angle is near zero; in which case any orientation

will generate identical amounts of electricity. For this reason, the insolation data provided by the

Electrical Contractors Association (2012) is limited to orientations of ±45˚ (SW and SE). Installations

that fall within these limits have to be further classified and need rounding to the nearest 5˚ value.

The optimum orientation for any installation is directly due South, unless external factors such as

shading are at play.

The power density incident on a PV module is at its maximum when perpendicular to the sun. The

relationships between panel inclination, orientation, and level of solar radiation are as follows;

𝑆𝑀𝑜𝑑𝑢𝑙𝑒 = 𝑆𝐼𝑛𝑐𝑖𝑑𝑒𝑛𝑡[cos(𝛼) sin(𝛽) cos(𝜓 − Θ) + sin(𝛼) cos(𝛽)]

Equation 1.5-9

Where SModule is the radiation incident on the module surface, SIncident is the radiation measured

perpendicular to the sun, ψ is the orientation angle of the module surface, Θ is the sun azimuth

angle (changing through the day), with α and β defined in Figure 1.5-8. When ψ = Θ, and α + β = 90˚

the module is directly facing the sun and generation will be at its maximum.


33

Figure 1.5-8: Variables definition solar radiation equations on a tilted surface.

Systems exist that are able to move the PV modules throughout daylight hours and follow the

movement of the sun in the sky; increasing the energy density of the sunlight incident on the

module surface. Tracking systems can either be vertical axis, inclined axis, two-axis tracker; the

layout of these systems can be seen in Figure 1.5-9. These systems increase installation and

operating costs for relatively little benefits to rooftop arrays. However, for utility scale installations

tracker systems have been known to increase annual received solar radiation by as much as 21%-

30%, depending upon the type of tracking used (Helwa, Bahgat, El Shafee, & El Shenawy, 2000). One

potential future study could include tracking system options and the potential increase in

generation. Unfortunately, there is little research into the precise increase in solar insolation these

systems induce for UK latitudes.

Figure 1.5-9: Different tracking systems employed by free mounted PV solar arrays (adapted from Sedona Solar Technology, 2016).


34

Figure 1.5-10 illustrates the expected performance for non-tracking systems of different inclinations

and orientations for a location in the centre of the UK. However, MCS specify that Figure 1.5-10 is

merely indicative of a single location and must not be used for performance assessment calculations.

Instead, the kWh/kWp charts must be referred to for generation calculations; an example of a zone

irradiance chart can be seen in Figure 6.3-2.

Figure 1.5-10: Effect of array orientation and inclination on system performance for a property in the middle of UK. Values are representative of a percentage of the maximum yield (South facing, 35˚

inclination). All values are taken from Electrical Contractors Association (2012).

1.5.2 Shading

Shading, or non-uniform illumination, is a major problem when generating solar energy. Causes of

shading such as clouds, fog, trees, and surrounding buildings and structures occur at a significant

distance from the panels themselves. However dirt, leaves, bird droppings, and water droplets can

also cause high levels of shading upon the panel surface itself (Lui, Pang, & Cheng, 2010).

For a PV module, the shaded cells will not produce a current equal to the unshaded cells.

Ramabadram & Mathur (2009) state that “there is a substantial power loss due to non-uniform

illumination of a series string care should be taken to see that all the cells connected in series receive

the same illumination under different patterns of shading”. As most modules connect cells in series,

the current through each cell must be the same. Forcing a higher current through the short circuit

current cells will induce a negative voltage. This causes a net voltage loss as the short circuit cells

essentially act as a load. This load is transformed into heat energy, causing hot spots on the module

surface. For this reason, a PV module current is limited by the current of the cell with the least

illumination without the use of bypass diodes (Karatepe, Boztepe, & Colak, 2007). The potential

effects of hot-spot damage are listed in Table 1.5-1.

N NW W SW S SE E NE N

-18

0

-16

5

-15

0

-13

5

-12

0

-10

5

-90

-75

-60

-45

-30

-15 0 15

30

45

60

75

90

10

5

12

0

13

5

15

0

16

5

18

0

90 25 27 32 37 40 44 50 57 60 66 70 70 70 70 70 66 60 57 50 44 40 37 32 27 25

80 28 29 34 38 45 50 56 61 68 73 78 79 78 79 78 73 68 61 56 50 45 38 34 29 28

70 32 34 37 42 48 55 66 69 75 80 85 87 87 87 85 80 75 69 66 55 48 42 37 34 32

60 37 38 41 48 54 62 69 75 80 85 91 92 93 92 91 85 80 75 69 62 54 48 41 38 37

50 42 45 48 52 59 68 72 78 85 91 95 96 97 96 95 91 85 78 72 68 59 52 48 45 42

45 46 49 52 56 62 70 74 80 87 92 95 97 98 97 95 92 87 80 74 70 62 56 52 49 46

40 50 52 55 60 65 73 76 82 88 93 96 98 99 98 96 93 88 82 76 73 65 60 55 52 50

35 55 57 59 65 67 75 79 85 90 94 97 99 100 99 97 94 90 85 79 75 67 65 59 57 55

30 60 60 61 67 70 78 80 85 91 94 96 98 98 98 96 94 91 85 80 78 70 67 61 60 60

20 70 70 71 73 75 82 84 86 91 92 95 96 96 96 95 92 91 86 84 82 75 73 71 70 70

10 78 78 79 80 82 85 85 88 90 91 92 92 93 92 92 91 90 88 85 85 82 80 79 78 78

0 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85 85

Incl

inat

ion

(˚)

Orientation (˚)


35

Table 1.5-1: The effects of varying hot-spot temperatures (Abdalla, 2013)

Hot-spot Temperature Damage to Array Potential Consequence

< 150˚C No damage No lasting effects

150˚C - 170˚C Encapsulation melting Delamination of the heat conducting material

170˚C - 200˚C Back sheet deterioration Reduction of the electrical isolation

> 200˚C P-N junction is destroyed Complete loss of PV operation

Bypass diodes are used to prevent any damage that may be caused by module hot spots. Typically

connected in parallel with opposite polarity to groups of cells (for cost reduction purposes), bypass

diodes will be reverse biased under normal operation. Under shaded conditions, however, the

bypass diode conducts, allowing the current to flow to the external circuit. This limits the negative

voltage bias to the single diode drop, preventing voltage loss, hot spots, and damage.

Figure 1.5-11: Diffuse/Global irradiance ratio plot for UK latitude values taken from Muneer (2004).

For opaque objects, such as leaves, the output of a cell is directly proportional to the surface area of

the cell covered. Distant forms of shading will not completely remove the diffuse light hitting the

module surface, therefore when the shadow of a structure completely covers the module area, the

diffuse insolation will still induce a current from the diffuse radiation.

As level of distant shading affects the direct irradiance only, the assessment tool will reduce amount

length of day proportionally. For example, 20% shading will reduce the length of day to 80% of its

maximum, where the remaining 20% will generate levels associated with diffuse radiation only.

y = 0.0101x + 0.0513 R² = 82%

0.45

0.5

0.55

0.6

0.65

0.7

46 48 50 52 54 56 58 60 62

Dif

fuse

/Glo

bal

Irra

dia

nce

Rat

io

Latitude (˚)


36

Table 6.3-3 gives diffuse/global irradiance ratio readings taken from stations across the UK. The

values range from 49˚ - 60˚ latitude. A linear relationship was identified and found to have an R2

value of 82% (see Figure 1.5-11), this was deemed suitable for calculations in EG’s assessment tool.

Other calculators that factor shading levels into their calculations have to state the level of accuracy

for MCS approval (Electrical Contractors Association, 2012). Therefore, ± 18% was deemed

acceptable to carry over. The calculation used in the assessment tool is as follows;

𝐼𝑟𝑟𝑎𝑑𝑖𝑎𝑛𝑐𝑒 𝑝𝑜𝑠𝑡 𝑠ℎ𝑎𝑑𝑖𝑛𝑔 = 𝐷𝑖𝑓𝑓𝑢𝑠𝑒 𝐼𝑟𝑟𝑎𝑑𝑖𝑎𝑛𝑐𝑒 + (𝐺𝑙𝑜𝑏𝑎𝑙 𝐼𝑟𝑟𝑎𝑑𝑖𝑎𝑛𝑐𝑒 × 𝑆ℎ𝑎𝑑𝑖𝑛𝑔 𝐹𝑎𝑐𝑡𝑜𝑟)

Equation 1.5-10

When the latitude and total irradiance values have been specified for the property the D/G ratio can

be found from the relationship above. The total irradiance is then split between diffuse and global

irradiance depending of the D/G ratio identified. Once specified the global irradiance is then

multiplied by the shading factor and added to the diffuse irradiance to find the new yearly total

irradiance value for the location.

1.5.3 Weather Patterns

The UK is not famed for clear skies and sunny days. For this reason, cloud coverage plays a major

factor in the performance on silicon PV arrays. Cloud coverage acts in a similar fashion to shading, in

that it reduces irradiance to diffuse only levels. As the data comes in kWh/kWp units, the weather

patterns for each zone are already accounted for in MCS recordings. The kWh/kWp data is real

historical data from stations across the UK, undergoing changing weather patterns that have been

averaged out for the collection period. Nonetheless, coastal regions can average 4% less cloud

coverage than inland regions within the same MCS zone (NASA, 2016). Therefore, a potential future

study could include how much cloud coverage changes within the boundary limits of the 25 zones,

allowing the tool to draw information from, NASA tabulated, 10km × 10km insolation areas for

generation calculations.

Table 1.5-2: Cloud coverage data for two opposite boundary areas within zone 13 (NASA, 2016).

Location

% Cloud Coverage

Janu

ary

Febru

ary

March

Ap

ril

May

Jun

e

July

Au

gust

Septe

mb

er

Octo

ber

No

vemb

er

De

cemb

er

An

nu

al

Average

Powys Lat. 52.6

Lon. 3.4 72 78 76 74 74 79 74 72 72 78 70 69 74

Milford

Haven

Lat. 51.47

Lon. 5.1 69 73 73 66 64 71 70 67 65 75 72 71 70


37

1.5.4 Degradation

Most solar PV solution suppliers offer a 20 year guarantee, however modules will most likely survive

40 to 50 years in non-maritime locations. Nevertheless, over time the peak output power of a PV

module will gradually reduce due to degradation mechanisms. This may be caused by a decreased

adherence of contacts (caused by water vapour), metal migration through the p-n junction, or

deterioration of the anti-reflective coating. Cracked cells (caused by hail, thermal stresses, or ‘latent

cracks’ that appear over time) are also of concern and must be considered when calculating

maintenance costs and system losses during financial assessments. The mean normalised

degradation for silicon PV in the UK is -0.8% per year, and median is -0.5% per year (Taylor, Leloux,

Hall, Everard, Briggs, & Buckley, 2015). The mean value from this study will automatically be

selected for use in the assessment tool as a cumulative linear factor; however the user will be able

to specify their degradation factor between 0% - 3% if desired.

1.5.5 Equipment Losses

System losses are inherent to any energy generation system and result in the power delivered to the

electricity grid to be lower than the power produced by the PV modules. Whilst solar PV systems are

limited by the conversion efficiency of the module being used, this conversion efficiency defines a

maximum performance index if no other losses are considered in the system.

The nominal peak power is the power rating given by the manufacturer of the module or system. If

your modules were 100% efficient, 1 m2 is needed to get a peak power of 1kW, under STC.

However, due to system losses, a larger area is required for the equivalent peak power. If you have

10% efficient modules you need 10m2 to have a 1kWp system, and so on;

𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝑃𝑒𝑎𝑘 𝑃𝑜𝑤𝑒𝑟, 𝑘𝑊𝑝 = 𝐴 × 𝑃𝑅

Equation 1.5-11

Where A is the area of the module (m2), and PR is the performance ratio of the system (%).

Table 1.5-3: Losses and their typical magnitudes in a PV system.

Loss Magnitude Standard value

used in tool Reference

Inverter losses 4%-9% 8%

(PV & SOLAR EDT, 2016) DC cable losses 1%-3% 2%

AC cable losses 1%-3% 2%

Performance Index 85%-94% 88%

Temperature losses 5%-18% 12% (Lelouxa, Narvartea, &

Trebosc, 2012) Performance Ratio 67%-89% 85%


38

Most assessment tools assume a performance ratio of 85% when generating quotes, as this is

considered a conservative estimate for modern PV systems (Lelouxa, Narvartea, & Trebosc, 2012).

When temperature losses are also considered the remaining factor is called the performance ratio.

The losses leading to a diminished performance index and ratio are listed in Table 1.5-3.

A performance ratio of 85% will be used in standard calculations unless otherwise specified by the

user. This value has been named ‘system losses’ to avoid confusion with an available range of 1-

20%.


39

1.6 Life cycle of PV modules

1.6.1 Energy Pay-Back Time

The energy payback time (EPBT) is defined as the time taken for an energy generating system to

produce as much energy as was consumed during the manufacture and system lifetime. Therefore,

EPBT can be minimised by reducing energy requirements during manufacture & installation,

lowering operational energy required, or increasing the generation efficiency of the system. For a

more in depth description of the production of PV modules see Figure 6.2-1 in the appendix.

𝐸𝑃𝐵𝑇 =𝐸𝐼𝑁𝑃𝑈𝑇 + 𝐸𝐵𝑂𝑆

𝐸𝑂𝑈𝑇𝑃𝑈𝑇

Equation 1.6-1

Where EINPUT is all energy involved in the manufacture, transportation, installation, operation,

maintenance, and module decommissioning/recycling. EBOS is the energy requirement of the

balance of system (BOS) components (e.g. support structures, cabling, electronic and electrical

components, and inverters). EOUTPUT is the annual energy savings of the system.

From a review completed in 2000 by Blakers & Weber, the average amount of energy required for

the production and installation of a solar panel internals was 1060kWh/m2, with silicon wafers

production accounting for > 70% of this amount. Other than the actual solar cell, the aluminium

framing accounts for a large proportion of material production energy usage; frameless modules

required 198MJ/m2 less energy than standard modules in 1997 (Keoleian & Lewis, 1997). When the

balance of system energy is taken into consideration for Sydney (where an annual yield of

153kWh/m2.annum was assumed) this translated as an EPBT of 8.5 to 11.5 years. Figure 1.6-1

displays the breakdown of the major processes involved in PV module system delivery and their

equivalent payback time both for known 2000 values and estimated 2010 values.

As of 2013, the median total energy requirement for a framed crystalline Si PV module had fallen to

312kWh/m2 with the BOS energy now accounting for over 30% of the total energy requirements

(Peng, Lu, & Yang, 2013). For installations of similar solar insolation levels to Sydney, further

reducing the EPBT range to 1.5-2 years. With suppliers usually offering a 20 year lifetime guarantee,

each panel should generate enough energy to produce a minimum ten further panels during this

time period. However, the outlook for the UK is not as positive where a 3.6-5 year EPBT is predicted

due to the lower solar irradiance levels and higher engineering costs. If this trend were to continue,

EPBT of only 1-2 years could be common, even in low irradiance locations such as the UK (Peng, Lu,

& Yang, 2013).


40

A study by researchers from Holland and the USA (Fthenakis, Kim, & Alsema, 2008), which analyses

PV module production processes based on 2004-2006 data, found that 250kWh is required to

produce a 1m2 crystalline Si panel for utility scale installations. This value will be used for EPBT

calculations in the solar assessment tool.

Figure 1.6-1: EPBT for processes involved in the manufacture, installation, and operation of a crystalline Si solar module installed in Sydney in 2000 and estimated values in 2010 (adapted from

Blakers & Weber, 2000).

1.6.2 Cups of Tea Equivalent

The total electricity generated over the lifetime of the panels will be reported to the user in the

results print-out in kWh. This unit is rather abstract unless the user is aware of energy patterns and

electricity requirements. For this reason, alternative energy generation reports are to be included

also. One, UK centric, idea was to report the number of cups of tea the user could make with the

energy saved through the installed PV system. This would be a simple linear conversion, assuming

an energy requirement of 2000W to operate a kettle for 2 minutes (Sust-it, 2016). Obviously, the

embedded energy in the manufacture, processing, and transport of the tea leafs, milk, sugar, and

packaging will far outweigh this, but is not a cost concern to the end user, and will be ignored for this

calculation.

1 𝑘𝑊ℎ =1

2𝐶𝑢𝑝 𝑜𝑓 𝑇𝑒𝑎

Equation 1.6-2

0

1

2

3

4

Productionof MG-Si

Productionof EG-Si

Productionof Cz Silicon

CellFabrication

PanelAssembly

BOS

EPB

T (Y

ear

s)

2000 Levels

2010 Levels


41

1.6.3 Greenhouse Gas Equivalence

Disputably, the key driving factor for an industry wide adoption of solar technology is climate

change. Climate change agreements (e.g. Kyoto Protocol in 1997) limit the amount of CO2 and other

GHGs governments can emit before receiving sanctions and fines. The most recent UN summit, held

in Paris 2015, agreed upon reducing carbon output in order to limit global warming to “well below

2°C” (UNFCCC, 2015). For this reason, carbon taxes levy industry, aiming to make renewables, CCS,

or CCU more economically viable. In the latest budget, Chancellor George Osbourne announced a

carbon floor price freeze at £18 per tonne CO2 until 2020 (Javid, 2014). However, this is expected to

race up to over £100 per tonne CO2 by 2035 (Gauke, 2015).

Solar energy has life cycle GHG emission levels of 22g/kWh for solar thermal and 46g/kWh for PV

(solar insolation level of 1700kWh/m2 assumed), with levels forecasted to fall as low at 15g/kWh

(Alsema, Wild-Scholten, & Fthenakis, 2006). Figure 1.6-2 displays the difference in CO2 equivalent

emissions for solar energy versus conventional technologies. Even with full investment in CCS

technologies for conventional O&G power generation methods, solar energy emits substantially less

CO2 per kWh generated.

Figure 1.6-2: CO2 equivalent emission levels for different energy sources currently, including forecasted CCS levels (solar insolation = 1700 kWh/m2 assumed) (adapted from Working Group III,

2014).

CO2 equivalent savings is typically the value displayed in the results page of a solar calculator. This is

quantified as the savings from the offsetting of the burning of coal in a UK standard pulverised coal

power station. For the purposes of this report and subsequent assessment tool this was taken as ≈

490gCO2/kWh (Lenzen, 2008). However, when given in kg of CO2, the total savings can be hard to

0

200

400

600

800

1000

Coal BiomassCofired

Gas Utility PV Rooftop PV CSP

CO

2 E

mis

sio

n (

g.e

q./

kWh

) Current Median Levels

Forecasted CCS Potential

Current Solar Levels


42

visualise. For this reason, the tool will also display two alternative results relating the CO2 offset to

more tangible values (miles offset for average passenger vehicle and acres of forest equivalent).

1.6.4 Offset CO2 equivalent calculations

Offset CO2 in kg is a rather abstract value that may mean little to nothing to the user. Instead, this

will be quantified as the number of miles an average US passenger vehicle can travel before emitting

the same amount of CO2 offset by the solar module over its lifetime. The weighted average fuel

economy for light road vehicles in the US was 21.4 miles per gallon in 2011 (FHWA, 2013). For the

same year, the total CO2 emitted per gallon of gasoline burned was 8.89 kg. The calculation to

convert CO2 offset to number of miles offset was therefore;

𝑀𝑖𝑙𝑒𝑠 𝑂𝑓𝑓𝑠𝑒𝑡 =𝐶𝑂2 𝑜𝑓𝑓𝑠𝑒𝑡 (𝑘𝑔) × 24.1 (𝑚𝑖𝑙𝑒𝑠)

8.89 (𝑘𝑔)

Equation 1.6-3 (FHWA, 2013)

Trees act as a carbon store, hence the process of planting and allowing a forest to grow to maturity

is carbon reducing. Data on the net change in forest carbon stocks from EPA (2012) was used to

identify a linear relationship between the acres of forest and CO2 sequestered per annum;

𝐴𝑟𝑒𝑎 𝑜𝑓 𝐹𝑜𝑟𝑒𝑠𝑡 𝐸𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡 (𝑎𝑐𝑟𝑒) = 1220 (𝑘𝑔 𝐶𝑂2 𝑠𝑒𝑞𝑢𝑒𝑠𝑡𝑒𝑟𝑒𝑑 𝑝𝑒𝑟 𝑎𝑛𝑛𝑢𝑚)

Equation 1.6-4 (EPA, 2012)


43

1.7 Financial Assessment

Financial assessments for solar panel installations are typically completed by the panel selling

companies when delivering quotes for potential customers. Values that are almost always quoted

are lifetime profit and payback period (PBP). These values are estimated through the summation of

a combination of savings. The savings have been classified for the purpose of this report as self-

consumption, Feed-in Tariff, and export tariff.

1.7.1 Self-Consumption

In the UK, most energy companies assess the electrical usage of a property from regular meter

readings. The average domestic cost per kWh in the UK as of January 2016 is 15.4p/kWh (EG-Audit,

2016). This varies dramatically depending upon the consumer (industrial rates are lower than

domestic rates) and pricing structure. When solar panels are generating electricity, the meter will

reduce the amount being consumed by the property from the electricity supplier. This, in turn,

reduces the total usage between future meter readings. This is the case for most meters; however

some meters have been retrofitting in order operate with bi-directional functionality (where meters

can read how much electricity is taken from the grid and how much is export to the grid).

Multiplying the supply cost per kWh by the solar generated kWh consumed by the property gives the

total savings by self-consumption. Naturally, during low production hours and high demand periods,

the proportion of generated energy being consumed will be close to 100%, however during peak

sunlight hours this could be much smaller.

Figure 1.7-1: PES areas of the UK and their distributor code number (adapted from OFGEM, 2016).


44

Electricity prices per kWh range from 6.5p - 13.2p before VAT and extra charges, depending upon

the user’s distribution network operator (DNO) and tariff structure (see Table 1.7-1). Energy

distributors were formed upon the privatisation of the electricity market in 1990; this required the

UK to be split into 14 public electricity supplier areas (PES) that can be seen in Figure 1.7-1.

The PES areas do not align with the MCS irradiance zones, therefore the user will need to specify

their DNO for calculation. If the DNO is unknown, then the average UK unit cost will be applied as

standard.

Table 1.7-1: Domestic cost per unit consumed for each DNO as of January 2016 (Gardner, 2015; EG-Audit, 2016).

Region

Average 2

01

5 B

ill (£)

Un

it cost in

c. charge

s and

VA

T fro

m B

ill (p/kW

h)*

Average Unit cost before (p/kWh)**

All U

nits

Day U

nits

Nigh

t Un

its

We

ek D

ay Un

its

Even

ing an

d

We

eken

d U

nits

No

n W

eek D

ay Un

its

Off P

eak U

nits

East Midlands 562 14.8 10.4 11.3 6.5 10.9 10.1 9.1 6.6

Eastern 561 14.8 10.8 11.5 6.6 11.2 10.4 9.3 6.8

London 586 15.4 10.4 11.4 6.7 10.8 - 9.2 6.8

Merseyside & North Wales 628 16.5 12.2 13.2 6.9 13.0 12.1 - 6.9

North East 578 15.2 10.7 11.5 6.7 11.3 - 9.7 6.7

North Scotland 635 16.7 11.6 12.6 7.7 12.0 - 10.9 8.1

North West 592 15.6 11.5 11.5 6.6 11.5 10.6 9.5 6.7

South East 577 15.2 10.7 11.7 6.5 11.0 10.2 9.2 6.9

South Scotland 571 15.0 11.0 12.3 7.1 11.4 - 10.2 7.6

South Wales 618 16.3 11.5 12.3 6.9 11.7 - 10.2 6.9

South West 613 16.1 11.6 126 6.7 12.0 11.3 10.2 6.7

Southern 576 15.2 10.9 11.8 6.8 11.5 10.7 9.4 6.8

West Midlands 579 15.2 10.9 11.7 6.6 10.7 - 10.5 6.9

Yorkshire 571 15.0 10.6 11.5 7.3 11.3 10.2 9.4 7.0

UK 584 15.4 11.1 11.9 6.8 11.5 10.7 9.8 7.0

* All bills are calculated assuming an annual consumption of 3,800 kWh. Unit costs reflect the prices of all suppliers and

include standing charges. Figures are inclusive of VAT. Unit costs are calculated by dividing the bills shown by the relevant

consumption levels.

** All values taken from EG-Audit (2016). Values are correct for DNO quote prices as of 1st

April 2016.

In total, there are 97 pricing structures DNO’s offer to electricity customers (EG-Audit Ltd., 2015);

this is before including any bespoke offers distributors regularly offer to larger consumers. The


45

purpose of pricing structures is to allow the consumer to pay more or less throughout the day,

depending upon their typical energy usage. For example, if a consumer feels that they use a

disproportionately high number of units during ‘evening and weekend’ hours, it would be in their

interest to select a pricing structure that charges less for units during this period.

As tariffing information for differing electricity packages is available to EG-Audit a potential future

study could include a suggested tariff swap after solar panel installation. An initial attempt at tariff

suggestion has been created and included in EG’s final tool, comparing four major pricing structures,

the results of which are included in section 2.4.

1.7.2 Export Tariff

The export tariff is a government legislated, privately funded, scheme that tracks and pays

microgeneration systems for any excess energy that is fed into the grid rather than used for self-

consumption. Introduced the same time as the FiT scheme, there have been many changes to the

‘floor price’ energy companies must offer their customers to purchase and resell this excess energy

(see Table 1.7-2). This gives the power to customers to negotiate higher export tariffs with their

energy company, to maximise their solar generation savings. This tariff is seen as mutually beneficial

for both customer and energy supplier. The energy generated by the solar panels that would

otherwise have gone to waste (unless some form of energy storage system has been installed in

parallel) is purchased, then resold, by the energy supplier for an average profit of ≈ 10p/kWh.

Table 1.7-2: UK ‘floor price’ export tariff for solar installations since introduction (Feed-in Tariffs, 2016).

Qualification Date Export Tariff (p/kWh)

April 2010 - April 2011 3.00

April 2011 - April 2012 3.10

April 2012 - August 2012 3.20

August 2012 - April 2013 4.50

April 2013 - April 2014 4.64

April 2014 - April 2015 4.77

April 2015 - Present 4.85

Unless SMART meters have been retro-fitted to the property, 50% of all energy generated is

assumed to be fed into the grid. It is the aim of the UK energy sector to roll out SMART meter

technology to all properties by 2020 (Samarakoon, Ekanayake, & Jenkins, 2011). If a property

believes it will export much more than 50% of the energy is generates, a SMART meter can be fitted

upon request. For installations greater in size than 30kWp an export meter is a contractual


46

requirement. The tool will automatically swap from applying a 50% export assumption, to the

calculated export ratio when a system size greater than 30 kWp is being assessed.

1.7.3 Feed-in Tariff

The FiT is a government scheme rolled out in early 2010, replacing government grants as the key

financial incentive for renewable microgeneration systems. Historically, the FiT rate has been

reviewed and changed significantly on three occasions (see Table 1.7-3 for historical FiT rates);

Table 1.7-3: UK FiT rate for solar installations since introduction (Feed-in Tariffs, 2016).

System Size

(kW)

April 2010 -

August 2011

August 2011 -

April 2012

April 2012 -

February 2016

February 2016 -

Present

Low High Low High Low High Low High

< 4 36.1 41.3 37.8 43.3 ≈ 11.7 ≈ 12.8 4.39

4 < 10 ≈ 36.8 37.8 ≈ 10.5 ≈ 11.7 4.39

10 < 50 ≈ 32.2 32.9 ≈ 10.5 ≈ 11.7 4.59

50 < 100 ≈ 32.2 19.0 ≈ 8.6 ≈ 9.6

2.70

100 < 250 ≈ 30.0 ≈ 17.0 2.70

> 250 ≈ 30.0 8.5 ≈ 6.2 2.27

* All values are given in p/kWh generated and taken from www.fitariffs.co.uk.

The payment period, since FiT introduction, has always been for 20 years. This allows early adopters

to continue claiming at least 36.1 p/kWh of electricity generated for a further 13 years. FiT is

assessed and adjusted in conjunction with the rapidly decreasing costs of solar panel installation and

the national economy, offering similar returns on investment from mid-2010 to early 2016.

However, the most recent adjustments have significantly removed the financial appeal of

installation; despite slight increases in the export tariff (see Figure 2.1-5). A property will be

classified into the low or high rate bands depending upon the buildings EPC rating. An EPC rating of

D or higher qualifies the property for the high FiT rate; this encourages property owners to install

insulation or other such measures to reduce energy waste. MCS approval is required for a property

to qualify for any FiT payment; encouraging the use of government certified installation companies

and guaranteeing that the installed technology meets the most recently legislated safety and

environmental regulations.

FiT, much like the export rate, also assumes that 50% of all generated electricity is fed into the grid

rather than consumed on-site, unless SMART meters have been fitted to the property. For example,

every 2kWh of electricity generated by solar panels, the customer will receive 4.39p if the

installation is of 4kW or smaller in size as of February 2016.


47

1.7.4 Annual Savings

The sum of the three saving factors described above equals the annual savings for PV systems.

Generally, savings are quoted over the course of 20 years. The export and FiT tariffs will remain a

flat rate from the point of contract; however the self-consumption savings will change over the

course of twenty years. Other calculators, most notably Solar Century, assume an annual 5%

increase in energy prices over this time period (a 165% increase in electricity price per kWh over the

tariff payment period). This value has been deemed a significant overestimation as energy prices

since 1996 have only seen a ≈ 76% increase over the same time period (Lucas H. , 2014). This

equates to an average ≈ 3% compounding annual increase that will be used as standard for the

assessment tool. A range of 0% - 6% will be available for selection if desired.

𝑆𝑒𝑙𝑓 𝐶𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 𝑆𝑎𝑣𝑖𝑛𝑔𝑠 = 𝑘𝑊ℎ 𝑆𝑎𝑣𝑒𝑑 × 𝐸𝑙𝑒𝑐𝑡𝑟𝑖𝑐𝑖𝑡𝑦 𝑃𝑟𝑖𝑐𝑒 𝑝𝑒𝑟 𝑘𝑊ℎ1.03 × 𝑌𝑒𝑎𝑟 𝑁𝑢𝑚𝑏𝑒𝑟

Equation 1.7-1

The ever increasing importance of self-consumption for the overall PBP may be the reason why

alternative calculators overstate electricity price inflation.

𝑇𝑜𝑡𝑎𝑙 𝑆𝑎𝑣𝑖𝑛𝑔𝑠 = (𝑆𝑒𝑙𝑓 𝐶𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 + 𝐸𝑥𝑝𝑜𝑟𝑡 + 𝐹𝑖𝑇) × 20 𝑦𝑒𝑎𝑟𝑠

Equation 1.7-2

1.7.5 Profit, PBP, & Return on Investment

When calculating profits, the initial capital outlay and annual maintenance cost must be considered.

Operating costs are assumed to equal zero for all intents and purposes for this report and the

corresponding assessment tool.

𝑇𝑜𝑡𝑎𝑙 𝑃𝑟𝑜𝑓𝑖𝑡𝑠 = 𝑇𝑜𝑡𝑎𝑙 𝑆𝑎𝑣𝑖𝑛𝑔𝑠 − (𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝐶𝑎𝑝𝑡𝑖𝑎𝑙 𝑂𝑢𝑡𝑙𝑎𝑦 + 𝑀𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 𝐶𝑜𝑠𝑡𝑠)

Equation 1.7-3

Maintenance costs are assumed to be 1% of initial capital outlay per annum (Campoccia, Dusonchet,

Telaretti, & Zizzo, 2009). However, the user may select anything in the range of 0%-3% if desired

(insurance potentially allows a maintenance cost of 0%).

The initial capital costs have been deemed linearly dependent upon installation size in kWp (see

Figure 1.7-2), the R2 value for small scale PV installations is 98.6%. This relationship will be used to

calculate the anticipated system quote price. However, if the user has received a specific quote

price, they are free to enter this into the tool in place of the estimated value.


48

Figure 1.7-2: Average capital cost for increasing installation size in UK (adapted from The GreenAge, 2016)

Other costs such as insurance and planning permission will not be factored into the final profits

calculation, as these are considered secondary costs. One significant factor that has been seemingly

overlooked in alternative calculators is the inverter replacement cost. The typical lifespan of an

inverter is 10 years, and as such will need to be replaced at least once during the guaranteed PV

lifetime. The average cost when replacing an inverter as of 2016 is £1,000. This cost will be included

as an optional extra in the final assessment tool (The GreenAge, 2016).

The financial PBP is defined as the time taken for the cumulative savings to exceed the initial capital

cost and any maintenance costs to that point in time. The PBP will be reported in years to the

nearest decimal place in the final assessment tool.

𝑃𝐵𝑃 =𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐶𝑜𝑠𝑡𝑠 + 𝑅𝑒𝑝𝑙𝑎𝑐𝑒𝑚𝑒𝑛𝑡 𝐼𝑛𝑣𝑒𝑟𝑡𝑒𝑟 + (20 × 𝐴𝑛𝑛𝑢𝑎𝑙 𝑀𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 𝐶𝑜𝑠𝑡𝑠)

𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑛𝑛𝑢𝑎𝑙 𝑆𝑎𝑣𝑖𝑛𝑔𝑠

Equation 1.7-4

The return on investment (ROI) is typically presented as a percentage so the user can evaluate the

efficiency of the investment compared to other potential investment vehicles (ISAs, or stocks).

Investments of a comparable magnitude are listed in Table 1.7-4.


49

The equation for ROI is given below;

20 𝑦𝑒𝑎𝑟 𝑅𝑂𝐼 =𝑁𝑒𝑡 𝐼𝑛𝑐𝑜𝑚𝑒

𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡=

𝑇𝑜𝑡𝑎𝑙 𝑃𝑟𝑜𝑓𝑖𝑡𝑠

𝐶𝑎𝑝𝑡𝑖𝑎𝑙 𝑂𝑢𝑡𝑙𝑎𝑦 𝐴𝑛𝑛𝑢𝑎𝑙 𝑅𝑂𝐼 =

20 𝑦𝑒𝑎𝑟 𝑅𝑂𝐼

20

Equation 1.7-5

Annual ROI will be the result presented in the assessment tool, as this is typically what is used in

alternative calculators and other investment formats.

Table 1.7-4: Investments of comparable magnitude to solar PV capital outlay and their average ROIs.

Investment Type Investment

Limit Annual ROI 20 year ROI

Minimum

investment period

Property (1996-2016) N/A ≈ 8% ≈ 160% Buyer dependant

ISA £15,000 ≈ 3% ≈ 80% Two year periods

Savings Account N/A ≈ 1% ≈ 22% N/A

FTSE 100 (1996-2016) N/A ≈ 3.5% ≈ 70% N/A

Solar PV (pre Feb 2016) N/A

kWp dependent

Quotes range from

3.5% - 8.5%

70% - 170% 20 years

Solar PV (post Feb 2016) N/A

kWp dependent

Quotes range from

1.5% - 3.5%

30% - 70% 20 years

* All values for calculation taken from current investment rates, or historical market data. Solar PV values taken from

alternative calculators. ROI values calculated using Equation 1.7-5.


50

1.8 Alternative Calculators

Eight leading online calculators have been selected for testing and review. Within the sample, there

are three not for profit government initiatives, two primary solar solution providers, and three

secondary solar solution comparison providers. Such an array of sources was selected in order to

gauge differing external biases that may be causing overestimated or conservative quotes. The

opposing natures of the sources reflect a range of appraisal bias a consumer may encounter.

1.8.1 Calculators Selected

The Energy Saving Trust is a not for profit organisation funded by the government and the private

sector. Private investors such as Thames Water and P&G will not be of influence to the results as the

respective industries do not align with solar or the energy sector. The Energy Saving Trust is a social

enterprise with charitable aspects (Energy Saving Trust, 2016). Founded in 1992, after the global

Earth summit, its primary focus is consumers and UK households.

National Renewable Energy Laboratory (NREL) is also a not for profit, government funded,

contractor operated research and development facility based in Colorado, USA. The focus of NREL

ranges across all renewables and energy efficiencies; however their primary motivation is PV

development and deployment. Their calculator should not be influenced by commercial reasons,

but considerable external bias from private laboratory funding may affect results (Plunkett, 2011).

The Centre for Alternative Technologies (CAT) is an education and visitor centre demonstrating

practical solutions for sustainability, founded in 1973. They are funded by proceeds from the visitor

centre, courses, shops, mail order, publications, and by operating as a renewable energy

consultancy. Other than the need to stay in operation, there is little influencing their practices

employed onsite and online, helping them to abide by their mission statement in the “search for

globally sustainable, whole and ecologically sound technologies and ways of life” (Centre for

Alternative Technologies, 2016).

Solar Guide is a secondary installation comparison website for consumers interested in purchasing

solar panels. Solar Panels UK are a Renewable Energy Association (REA) and Solar Trade Association

(STA) affiliated company with a price comparison site that specialises in solar power. Of all the

tested calculators, Solar Panels UK seemed the least professional and targeted towards highlighting

financial benefits rather than the energy saving aspects of solar panels (see section 2.1 for a

complete review). Both mentioned companies receive a commission fee for successful connections

made through their sites, and as such it is in their interest to convince site users of the benefits of

solar technologies (Solar Guide, 2015).


51

The ECOEXPERTS claim to be the UK’s no. 1 solar panel comparison website as well as possessing a

calculator that “gives you the most accurate results available on the internet.” (The ECOEXPERTS,

2016). Since their founding, the ECOEXPERTS have been involved in over 25% of UK solar

installations. The ECOEXPERTS will be the highest profile company in the UK for review.

Founded in 1998, Solar Century is one of the leading primary solar solution providers in the UK (Solar

Century Ltd, 2016). Collecting many awards and accolades over the years, their calculator has a

reputation as one of the most reliable and accurate. Profits relying on customers continuing to

purchase their solar panels, however, may influence results and result presentation methods.

SolarWorld is a German based company, with a UK subsidiary, dedicated to the manufacture and

wholesale of PV products worldwide. Shell divested the majority of its solar interests to SolarWorld

in 2006, leading to SolarWorld becoming the largest PV providers in North America (Renewable

Energy World, 2006). As with other suppliers, the results for SolarWorld UK’s calculator may be

affected by internal biases.

1.8.2 Review Method

An example property (Figure 1.8-1) was used to generate quotes from the alternative solar

calculators. The position and height of the neighbouring property led to the ‘light shading’ factor,

specified, when requested, in the alternative calculators. The selected roof is SE facing, with a pitch

of ≈ 45°, the total roof area is ≈ 40m2.

Figure 1.8-1: Property selected as ‘standard’ to compare assessment results from alternative calculators. Roof selected for PV installation has been highlighted (adapted from Google Maps).


52

The input values selected for the example property are as seen in Table 1.8-1;

Table 1.8-1: UK average values used to generate quotes for alternative calculator.

Property Value Unit Reason for selection

Location CH2 4DW N/A Based in Chester, North West

Inclination 45 ° Typical roof slope for UK new builds

Orientation South East N/A Median direction for potential installations

Installation size 4 kW Typical installation size for ‘domestic use’

(Muhammad-Sukki, et al., 2013)

Area available 40 m2 Equal to ≈ 4kW panel set up

Usage ratio estimate 50 % Assumed value if no option given

EPC rating D N/A Average UK EPC rating

Installation quote 7,000 £ Average 4 kW UK quote price

System losses 15 % Typical solar panel losses after year 1

Annual elec. usage 3,800 kWh Average household usage (Gardner, 2015)

Electricity rate 15.47 p/kWh UK national average 2015 (Lucas H. , 2015)

Monthly elec. bill 50 £ Calculated from above values

Export tariff 4.64 p/kWh UK national average 2015 (Renewable Energy

Consumer Code, 2015)

FiT 11.8 p/kWh UK national average 2015 (Renewable Energy

Consumer Code, 2015)

Shading < 20 % Considered ‘partial shading’

Module type Silicon PV N/A Most common solar panel installed

This is an overview of all the factors requested by the alternative calculators selected for review. For

a fair comparison, these values will be applied to the final tool developed for EG. The results from

each calculator will be compared against each other and against EG’s final tool.

Key Used throughout Alternative Calculator Review:

● Primary solar solution providers ● Solar Century & Solar Panels UK

● Secondary solar solution providers ● Solar World UK, The ECOEXPERTS, & Solar World UK

● Not-for-profit organisations ● NREL, CAT, & Energy Saving Trust


53

2.0 Results and Analysis

All results and figures included in this section were ascertained from the analysis of alternative

calculators or from the troubleshooting of scenarios in EG’s final assessment tool. See abstract of

this report for links to the documents used to amass the following results.

2.1 Analysis of Alternative Calculators

2.1.1 Review of Input Factors

Each alternative calculator required a different set of variables to be defined for property

assessment. Table 2.1-1 lists the factors requested by each calculator, and has ranked the

calculators according to the level of system definition required (the factors have also been ranked

according to the frequency of their application). It should follow that an assessment has the

opportunity to be more accurate for an increasing levels of specification. This has been assumed as

the highly specified calculators will inherently employ fewer assumptions during quote generation.

Table 2.1-1: Rival solar calculators ranked for factors employed when generating quotes.

Solar guide allowed the user to define more variables than any other option, closely followed by the

Energy Saving Trust and NREL. In saying this, the majority of Solar Guide’s factors are initially hidden

assumptions; only by selected ‘advanced assessment’ can a user enter unique values for the financial

assessment variables. Despite the high specificity levels of the other not for profit organisations,

Locatio

n

Inclin

ation

Orie

ntatio

n

Are

a available

(m2)*

Installatio

n Size

(kW)*

Usage

Ratio

Estimate

EPC

rating

Installatio

n q

uo

te

System

Losses

Electricity U

sage

Electricity rate

Expo

rt tariff

Shad

ing

Feed

in T

ariff

Mo

du

le T

ype

Pro

pe

rty Type

Total Sco

re /16

Solar Guide 10

Energy Saving Trust 8

NREL 8

Solar Century 7

Solar Panels UK 7

The ECOEXPERTS 6

Solar World UK 4

CAT 4

Frequency used /8 8 8 8 4 4 4 3 3 2 2 2 2 2 1 1 0

* Both factors define the same variable (system nominal peak power) via separate calculation methods. Technically this variable was required as frequently as location, inclination, and orientation.


54

CAT required the fewest factors for definition. CAT themselves acknowledge that the calculator is

only capable of creating a “rough estimate” and that the results should “never” be the only basis for

installation (Centre for Alternative Technologies, 2016). Interestingly, the two most financially

driven companies (Solar Century and Solar Panels UK) only allow the user to define three financially

bases factors between them; therefore assumptions will need to be employed that may lead to

inaccurate financial results.

Location, inclination, and orientation are factors consistently requested when generating any quote

(MCS requirement). Peak system generation was requested in every case as well, either as area

available or installation size; as they both define this variable via separate calculation methods.

Every other factor was assumed in at least one of the alternative calculators, with alarmingly few

calculators requesting the level of shading for the assessed property. Experiments by Ubisse &

Sebitosi have shown that even when relatively few cells are partially shaded, output can drop to as

low as 45% of peak performance.

Figure 2.1-1: Layout of EG’s solar assessment input page.

The final tool for EG retains the ability to define all of the above mentioned inputs, barring from

‘module type’. The reason for this oversight is that all research into the irradiance variables was

Variable Entry Carried Value

Postcode CH2 4DW -

Year of Installation 2016 -

EPC Rating C -

Panel Inclination (degrees from Horizontal) 45 -

Panel Orientation (degrees from South) 25 -

Area Classification Countryside 845

Shading Factor Light Shading (10%-20%) 765.3656583

Assumed System Losses 15% 650.5608095

Degradation Rate (per annum) 0.5% -

Nominal Peak Power of System (kWp) 4 -

Area Required (m2) 44.3 -

Typical Annual Electricity Usage (kWh/annum) 3800 -

Property Type Social -

April 2015 - Present -

4.85 -

February 2016 - Present -

4.39 -

Electricity Distributor UK Average -

Average Electricity Cost (p/kWh) 15.4 -

Installation Cost (£) 6170.8 -

Maintenance Costs (% Installation Cost per annum) 1.0% -

Will Inverter need Replacing? Yes

Lifetime of System (years) 20 -

Inflation over Time Period (%) 3.0% -

Automatic Meter Readings? Non-AMR

FiT Rate (p/kWh)

Export Tariff (p/kWh)


55

completed for an average silicon solar cell; rather than industry specific modules. Furthermore, EG’s

tool has factored in novel features such as;

Table 2.1-2: Novel features included in EG’s assessment tool.

User defined inputs Variable assumptions from ‘standard’ inputs

Area Classification (pollution level control) System Losses

Installation Date FiT and Export Rates Degradation Rate

DNO Selection Maintenance Costs

Inverter Replacement Electricity Pricing Inflation Rate

Property Classification -

This equates to a score of 19 for EG’s tool in terms of factors required; almost doubling the system

specification level of the current leader. One issue of this is that for the typical user many of these

factors may be unknown or misunderstood. To avert this, ‘standard’ values have been specified

where deemed necessary, as can be seen in Figure 2.1-1. Brief explanations have been set to pop up

when the user is defining an input, further reducing confusion during input definition. In most

circumstances, one of EG’s trained auditors will be completing the assessment on behalf of the

customer, however, a user guide has also been written for the assessment tool. This user guide can

be found by following the link included in the abstract to this report.

Significantly, the calculators analysed in this report have not updated the FiT and export rates to

match the post February 2016 values, as of research completed in March 2016. The impacts of this

oversight can be seen in Figure 2.1-5.

2.1.2 Review of Results Presentation

Each calculator employed different methods for presenting their results to the user. Moreover, the

type of results utilised varied considerably between the respective calculator types. The results

were classified either as ‘financial information’ or ‘environmental information’ for analysis purposes.

It was anticipated that the more financially driven companies would present their results in a

manner that positively reflected the monetary benefits of PV installation. Whereas, the not-for-

profit organisations were anticipated to highlight the environmental and financial benefits more

evenly.

The information was either presented in numeric or graphical format (see section 6.4 for all

instances where graphs were used to present data to the user). In either case the result value was

noted or interpreted for further analysis in Table 2.1-5. Table 2.1-3 and Figure 2.1-2 outline the

results presented; ranking each calculator from the most detailed assessment to the least. Each

result has been further ranked according to their frequency of presentation.


56

Table 2.1-3: Rival solar calculators ranked for displayed results for generated quotes.

Financial Information Environmental

Information

Self-con

sum

ptio

n savin

gs

FiT P

ayme

nt

Expo

rt Paym

en

t

Total P

rofit

Finan

cial breakd

ow

n grap

h

Finan

cial Payb

ack Pe

riod

Re

turn

on

Investm

ent

Pan

el recom

men

datio

n

CO

2 foo

tprin

t red

uctio

n

Total e

nergy gen

erated

Assu

me

d so

lar irradian

ce

Energy gen

eration

graph

Energy P

ayback P

erio

d

Total sco

re /13

CAT 9

Solar Guide 8

Energy Saving Trust 6

Solar World UK 6

Solar Panels UK 6

The ECOEXPERTS 4

Solar Century 4

NREL 3

Frequency used /8 7 6 6 5 4 3 2 1 4 4 2 1 1

Figure 2.1-2: Number of results displayed by rival calculators including discernible information from graphs (see section 6.4), ranked in descending level of detail.


57

On the whole, the results displayed by the primary solar solution providers were not as in depth or

detailed when compared to other reviewed calculators (ranking 5th and 7th out of eight for the

number of financial results, and 6th and 7th out of eight for the number of environmental results; see

Figure 2.1-2). Solar Panels UK fared slightly better than Solar Century; however the margin is only

two results. It will be interesting to see how the primary solar solution providers modify their result

presentation methods once their calculators are altered to take account of the lower FiT rate.

The secondary solar solution provider Solar Guide exhibits a financial result bias, detailing six results

of this nature versus one environmentally based result. A similar trend can be observed to a lesser

extent for Solar World UK, but as “one of the leading environmental information providers” it would

be unwarranted to attribute this to financial motivated partialities. The ECOEXPERTS proved

incredibly uninformative with only four results presented overall, all of which were financial in

nature; their operating business title can perhaps be considered a misnomer after this review.

Barring NREL, the not-for-profit organisations fared well in terms of overall ranking (ranking no lower

than 1st and 3rd of eight; as can be seen in Table 2.1-3). Despite this, the not-for-profit calculators

also consistently displayed more financial information than environmental information. Even

though CAT required the least system specification, it presented the most detailed results of any of

the assessed calculators. It should be examined why CAT feel that they can derive, and exhibit, more

conclusions than Solar Guide (whose system specification require over three times as many input

factors). It should be noted that the Energy Saving Trust presents the most uniform distribution of

financial and environmental results, as well as being one of the most specified assessments (see

Table 2.1-1). Overall, this is the form of result presentation that EG’s tool attempts to emulate.

Overall, financial information was presented 300% more frequently than environmental information.

Clearly, these companies believe that their users are more interested in the financial aspects of PV

installation rather than the environmental benefits. If this inference is true, the dramatic reduction

in the FiT tariff from February 2016 may have a considerable impact upon the result presentation

and calculation methods employed by these calculators in the future. Going forward, a repeat

review should be completed following any updates for the assessed calculators.

EG’s assessment tool will generate a print-out for the user to consider and compare against

alternative quotes received from alternative solar solution providers (see Figure 2.1-3). The print-

out contains 12 out of the 13 results used by the alternative calculators (panel recommendation

could not be included due to research limitations). Additionally, the print-out has included the

following factors novel to the assessed calculators;


58

Table 2.1-4: Novel results presented in EG’s assessment tool.

Financial Results Environmental Results

Actual Export/Self-consumption Ratio Energy Pay Back Time

Annual Electricity Usage Trend Daily Generation Curve

Daily Electricity Usage Trend Cups of Tea Equivalent

- Acres of Forest Equivalent

- Passenger Vehicle Miles Offset

This increases the number of results EG presents to 20; over double the level of detail presented by

CAT. Figure 2.1-3 demonstrates the presentation styles employed by EG’s results print-out. For a

full print-out sheet example see section 6.5. The ratio of financial to environmental information

displayed in EG’s print-out is 1:1. This ratio is representative of EG’s stance as an unbiased and

impartial party in the solar industry. In terms of calculation transparency, the print-out includes all

assumptions made for the financial assessment. Whilst a link to this report will be made available

for anyone interested in the methods used to calculate total solar generation for the evaluated

system.

Figure 2.1-3: Novel result presentation styles (images taken from EG’s assessment tool print-outs).


59

2.1.1 Numerical Results Comparison

The numerical results from the alternative calculators have been compiled and presented in Table

2.1-5. Where a result has been deemed anomalous it has been highlighted and italicised. These

values were not carried forward for average or standard error calculations; therefore, they will not

affect comparison against EG’s tool. Where possible, the average, standard deviation, and standard

error of these results have been included. In the case of EPBT, the one given value was taken as the

average for comparison.

Table 2.1-5: All numerical results gained from alternative calculator review.

An

nu

al Self-con

sum

ptio

n Savin

gs (£)

An

nu

al FiT Paym

en

t (£)

An

nu

al Expo

rt Paym

en

t (£)

Lifetime

Pro

fit (£)

PB

P (m

on

ths)

An

nu

al RO

I (%)

First Year G

en

eratio

n (kW

h)

CO

2 Equ

ivalen

t (kg/ann

um

)

EPB

T (m

on

ths)

CAT 226 389 73 6760 120 4.8 3010 1330 28

Solar Guide 196 392 76 5800 106 9.3 - - -

Energy Saving Trust 112 383 77 4000 - - 3184 1400 -

Solar World UK 298 398 80 8000 120 - 3310 1570 -

Solar Panels UK 258 381 50 10800 - - 3227 - -

The ECOEXPERTS 113 365 67 4160 - - - - -

Solar Century - - - 18000 72 15 - 1840 -

NREL 393 - - 7860 - - 3140 - -

Average 228 385 70.5 6760 115 7.0 3170 1540 28

Standard Deviation 101 11.4 11.0 2380 8.1 3.2 111 230 -

St.Dev. % of mean 44% 3.0% 16% 36% 7.0% 45% 3.5% 15% -

Standard Error* 38.0 4.67 4.48 900 4.6 2.3 49.7 110 -

* Calculated by dividing standard deviation by the square root of the number of readings.

As can be seen in Table 2.1-5, the results vary dramatically from one calculator to the next for some

results. This is most apparent for self-consumption savings, lifetime profits, and ROI, where the

standard deviation is approximately 40% of the mean value. As profit and ROI are subsequently


60

calculated from the self-consumption savings, it begs the question ‘how can this prediction be so

different despite identically specifying the systems’? This highlights the importance of accurate self-

consumption predictions on subsequent financial results; in turn the ability to better predict the

export/self-consumption ratio from the analysis of historical electricity usage. The annual

generation predictions are reasonably consistent, with a standard deviation only 3.5% of the mean.

Unfortunately, only one calculator applied this result for further analysis (CAT’s prediction of a 28

month EPBT). The three financial results from Solar Century were deemed anomalous for the

purposes of this report (an estimated profit 260% higher than the mean of the other seven sources).

These results had not been justified with a self-consumption, export, and FiT payment breakdown

payments, hence disregarded as unfounded.

The standard error for these results was calculated in order to give EG’s assessment results an

acceptable bracket to fall within. A practicable range of alternative calculators had been reviewed;

therefore the average results should give a theoretical indication of a ‘best knowledge’ prediction.

Any bias shown by one source should be counteracted by the seven other sources in the sample.

Contacting the developers of the reviewed assessment tools, to see the underlying calculation

methods could be completed in a potential future study. Taking a larger sample of calculators may

also reduce the standard error values calculated for the financial results.

Figure 2.1-4: Profits predicted by rival calculators ranked in ascending order.

0

4000

8000

12000

16000

20000

Life

tim

e P

rofi

t (£

)


61

Figure 2.1-4 exhibits the predicted profits for each reviewed calculator. It is rather telling that the

primary solar solution providers consistently supplied the largest profit predictions; separated by

substantial margin of £4,000 - £11,200 greater than the remaining sample average. The remainder

of the calculators supplied much more conservative estimates. The ECOEXPERTS, despite their

complete neglect of environmental information, predicted the second lowest profit for the modelled

system. Overall, the difference in predicted profits between the secondary solar solution providers

and the not-for-profit organisations is only 3.6%. This may be down to concern over reputational

damage, if the secondary solar solution providers are found to be consistently overestimating profits

compared to their competitors.

Figure 2.1-5 has been calculated from the revenue breakdown results from the alternative

calculators. By multiplying the FiT average value by the factor it will be changing as of February 2016

the RHS chart was plotted. With the loss of the higher FiT rates, as of February 2016, the export

tariff and self-consumption savings will make up a much larger portion of the estimated profit.

Falling from 56% of total returns to 31%, the FiT savings are no longer the major financial benefit for

installation (see Figure 2.1-5). Instead, the self-consumption savings become the majority

contributor to lifetime profits (increasing from 33.4% to 52.8% of total revenue).

Figure 2.1-5: Prospective savings breakdown before and after changes to the FiT rate (assumed 12.2p/kWh before January 2016 and 4.4p/kWh after February 2016).

33.4%

56.3%

10.3%

Revenue streams April 2012 - February 2016

Electricity Savings

FiT Payment

Export Payment

52.8%

30.9%

16.3%

Revenue streams February 2016 - Present

Electricity Savings

FiT Payment

Export Payment


62

This further highlights the importance of export/self-consumption ratio; as these two revenue

stream will compromise ≈ 69% of total future revenue streams. Despite increasing in p/kWh, the

proportion of total revenue that export payments consist of is substantially less than the other two

income methods (16.3% from February 2016 onwards).

2.2 Assessment Tool Results Comparison

2.2.1 Generation Profile

Figure 2.2-1 displays the calculated average generation curves for the different seasons of the year.

The peak generation expected during summer days is approximately double the equivalent peak on

a winter’s day. Average generation during spring days is, rather surprisingly, much more similar to a

typical day in summer, rather than an autumn day. These relationships highlight the significance of

optimising seasonal usage to minimise overall export ratios. Even if a property were able to

constructively change usage behaviours on a daily basis, it would be a far greater challenge to shift

electricity requirements from one season to another. This scenario confirms that the tool can

accurately model seasonal irradiance levels by comparison against ‘Tables of Temperature, Relative

Humidity, Precipitation and Sunshine for the World, Part III/ Europe and the Azores’ (UK

Meteorological Office, 1972).

Figure 2.2-1: Average seasonal generation trend for system specified in Table 1.8-1, data is taken from EG’s final assessment tool.

0

0.1

0.2

0.3

0.4

0.5

0.6

00:00 04:00 08:00 12:00 16:00 20:00

Hal

f-h

ou

rly

Ge

ne

rati

on

(kW

h)

Winter

Spring

Summer

Autumn


63

A potential future study could include a comparison of the alternative calculator results for varying

input factors, not the single specific system. Analysis graphs for the changes to individual inputs

would highlight the dependence upon these factors to estimated generation levels.

2.2.2 Actual Export Ratio

The export ratio was determined by plotting the expected electricity usage and generation levels for

every half hour throughout an average year. When generation exceeded the electricity usage, the

surplus was deemed to have been exported to the grid. Dividing this surplus by the total PV

generation gave the export ratios seen in Table 2.2-1. If anticipated export periods are known for a

given property, the occupants will be able to change their electricity consumption habits to minimise

the amount of electricity exported to the grid; in turn maximising self-consumption savings and

lifetime profits. As mentioned in section 1.7, SMART meters are only of benefit to the user when

more than 50% of generated electricity is exported to the grid. This basis was used to calculate the

kWp values given in Table 2.2-1.

Figure 2.2-2: Average daily usage versus generation curves for the commercial and residential properties for a 4kWp solar installation and electricity usage of 3,800kWh/annum.

Figure 2.2-2 displays the average usage versus generation curves for the commercial case study. A

period of net export can be seen at any point where the generation curve exceeds the usage curve.

For an average day of the year this does not occur at any point for the commercial property.

However, for the residential property the system will be exporting to the grid between the hours of

8AM and 4PM. This does not mean that the commercial properties export ratio is 0%, as what is

0

0.1

0.2

0.3

0.4

12:00 AM 06:00 AM 12:00 PM 06:00 PM

kWh

/hal

f h

ou

r

Commercial Usage

Residential Usage

4kWp Generation


64

shown is simply an ‘average’ day. There will be days (particularly during summer months) when

usage is less than average, and PV generation is higher than average, leading to net meter exports.

Table 2.2-1: Export ratio variance for EG supplied case studies.

Property Type Export

Ratio*

Export

Value

(£)

Equivalent Self-

consumption

Value (£)

Increase in Profit if

reduced to 0%

export ratio (£)

At what kWp would a

SMART meter be of

benefit?**

MCS Assumed 50% 60.0 191 2610 -

Commercial 5.5% 5.6 17.8 245 7.7

Educational 11.6% 11.2 35.6 490 7.1

Industrial 5.5% 5.6 17.8 245 7.7

Light Commercial 11.8% 11.3 35.9 490 7.0

Residential 21.2% 18.8 59.7 820 6.2

Social 12.0% 11.5 36.5 500 6.7

* Annual electricity usage of 3,800 kWh, an export rate of 4.85p/kWh, and an installed PV system capacity of 4 kWp was

assumed for all calculations. All other necessary system specifications are as seen in Table 1.8-1. Area classification of

‘town/village’ was used for calculation.

**Beneficial SMART meter kWp values are based off 3,800 kWh annual electricity usage.

Of the six case study properties supplied by EG, the residential property demonstrated the highest

export ratio with over a fifth, 21.2%, of all generated electricity transferred to the grid. The property

type that displayed the lowest export ratio was the commercial property with 5.5% export ratio. The

educational property, rather surprisingly, had only the 4th highest export ratio. This is contrary to

the prediction that the summer shut down period would boost the export ratio considerably higher

than the other property types.

Table 2.2-2 outlines the proportion of total annual export for each season. For all properties, export

during winter months accounted for less than 0.3% of the annual total. A combination of high

electricity usage and low PV generation levels agrees with this finding. For all case studies, barring

the residential property, roughly one third of total export takes place from March to May (spring);

the high proportional usage in March and April for the residential property seen in Figure 2.3-1

accounts for this difference. The summer months accounted for 53.5% of total annual export for all

properties analysed. For the educational property, June to August exported 59.1% of the annual

total; the highest for any property or season. The summer shut down period will contribute to this

higher than average proportional export level. Autumn exports saw the largest variance in

proportion between the different property types; ranging from 6.6% for the education property to

22.7% for the residential property. The residential property displays the most evenly distributed

seasonal export; most likely due to the uniform weekly trend seen in Figure 2.4-9.


65

Table 2.2-2: Seasonal export as a proportion of total annual export for each case study.

Season

Co

mm

ercial

Edu

cation

al

Ind

ustrial

Light C

om

me

rcial

Re

siden

tial

Social

Average

Winter 0.0% 0.1% 0.3% 0.3% 0.0% 0.0% 0.1% Spring 33.9% 34.1% 33.3% 35.5% 28.4% 34.3% 33.3%

Summer 57.1% 59.1% 50.5% 55.1% 48.9% 52.2% 53.8% Autumn 9.0% 6.6% 15.9% 9.2% 22.7% 13.5% 12.8%

* All values calculated for properties consuming an average 3,800kWh/annum, with 4kWp

PV system installed.

For reasons discussed in section 1.7, the export ratio is inversely proportional to lifetime profits. It

would be in the interest of the tenants of the properties that display high export ratios to change

their usage habits, in order to maximise profits. This can be done by holding off on energy intense

processes, such as using a washing machine, for periods of peak solar generation. Figure 2.2-3

displays the average seasonal export proportions for every property type visually. It reinforces how

important it is to be aware of electricity usage behaviour during summer months.

Figure 2.2-3: Seasonal export values ad a proportion of total annual export, averaged for all case study data supplied by EG.

Winter 0.1%

Spring 33.3%

Summer 53.8%

Autumn 12.8%

Average Seasonal Export as a Proportion of Total Annual Export


66

2.2.3 Results Comparison

In this section, the results gained from EG’s tool will be compared to the results gained from the

analysis of the alternative calculators. Table 2.2-3 displays the system specified for the novel

features in EG’s tool to gain the results seen in Table 2.2-4. All other system specifications are

identical to the values used to retrieve results from the alternative calculators in section 2.1.

Table 2.2-3: Additional inputs and the values selected for result retrieval from final assessment tool.

Variable Entry

Area Classification Countryside

Property Classification N/A*

Shading Factor Light Shading (10%-20%)

Degradation Rate (per annum) 0.5%

Export Tariff (p/kWh) April 2015 – Present

4.85

FiT Rate (p/kWh) April 2012 – February 2016

12.38

Electricity Distributor UK Average

Average Electricity Cost (p/kWh) 15.4

Annual Maintenance Costs (% Installation Cost) 1.0%

Will Inverter need Replacing? No

Inflation over Time Period (%) 3.0%

Table 2.2-4: All results for different property types relative to alternative calculator results.

Result

Alte

rnative

Calcu

lator R

esults*

50

% Self-co

nsu

mp

tion

**

Co

mm

ercial

Edu

cation

al

Ind

ustrial

Light C

om

me

rcial

Re

side

ntial

Social

EG To

ol A

verage

Min. Max.

Actual Export Ratio 50% 50% 5.5% 11.6% 5.5% 11.8% 21.2% 12.0% 11.3%

Self-consumption savings (£) 190 266 256 434 410 433 410 380 409 410

Export Payment (£) 66.0 75.0 60 60 60

FiT Payment (£) 380 389 297 297 297

Lifetime Profits (£) 5860 7660 6240 8380 7900 8370 7880 7250 7880 7940

PBP (months) 120 110 118 94 97 94 97 101 97 97

Annual ROI (%) 4.7 9.3 5.1 6.8 6.4 6.8 6.4 5.9 6.4 6.5

First Year Generation (kWh) 3120 3220 2540 2540 2540

CO2 Savings (kg/annum) 1430 1650 1240 1240 1240

EPBT (months) 28 71 71 71

*Standard error values from Table 2.1-5 were used to find the minimum and maximum value for each result.

**50% self-consumption assumption is what is used by most alternative calculators when financial assessment is completed.


67

Table 2.2-4 demonstrates that when a 50% export versus self-consumption ratio is assumed, the

majority of the financial results fall within the acceptable error margin of the assessment tool

sample. Nonetheless, the FiT payment is significantly lower than the minimum value deemed

acceptable in Table 2.1-5. This is a direct result of the lower first year generation result retrieved

from EG’s tool. EG’s tool could be deemed pessimistic, with a proposed 2540kWh generated

opposed to an alternative calculators average of 3170kWh. This value is lower than the acceptable

minimum because of the increased number of negatively correlated input variables included in EG’s

tool. Area classification, variable system losses, and degradation rate are all irradiance diminishing,

and typically overlooked by the alternative calculators. Additionally, EG’s shading factor calculation

is novel and may also be a reason why the annual generation prediction falls beneath the acceptable

minimum.

From the export ratio analysis, it was shown that the 50% assumption employed by al other

assessment tools was a considerable overestimation. For this reason, the self-consumption savings

calculated by EG’s tool for the case study properties are much higher than the sample average. As

self-consumption is the largest contributor to lifetime profits, it follows that the predicted profits,

and ROI results for the case study properties reflect this increase in annual revenue. Conversely, the

lower FiT payments predicted by EG are representative of the negatively correlated input variables

reducing predicted annual generation. The net change caused by these two revenue streams

culminates in predicted profits ranging from £6,240 for the commercial property to £8,370 for the

industrial property. The average profit predicted by EG’s assessment tool under the test conditions

is £7,940; £1,220 greater than the sample average and almost identical to the predictions made by

NREL and SolarWorld UK.

Figure 2.2-4: Environmental results for 4kWp system defined in Table 1.8-1.

Figure 2.2-4 gives an example of the unique results EG’s tool will display, and the presentation

method employed, for assessment print-outs. These values are dependent upon the system

location, orientation, inclination, and system size. 629,000 cups of tea equals the amount of

electricity generated over the system lifetime. Over 61,000 miles can be travelled in an average


68

passenger vehicle without exceeding the amount of CO2 saved by the generation levels from the

renewable PV source instead of a conventional coal fired station. This amount of CO2 is also equal to

the amount sequestered by 21 acres of forest in a year.

The predicted EPBT result for EG’s tool is far greater than CAT’s prediction (at 71 months, EG’s

prediction is 3 years and 7 months longer). CAT’s calculation method assumes that it takes “250kWh

of electricity to produce 1m2 of crystalline Si PV panel. Under typical UK conditions, a 1m2 panel will

produce around 100kWh electricity/annum, so it will take around 2.5 years to pay back the energy

cost of the panel” (CAT, 2016). In comparison, EG calculate EPBT via nominal peak power rating

rather than panel area. This method is far more accurate as it factors in system losses in place of an

arbitrary assumption for all installations.


69

2.3 Electricity Usage Trends

2.3.1 Annual Usage Trends

Figure 2.3-1 illustrates the monthly proportion of electricity used for each property type. This graph

was created from the analysis of the EG supplied half hourly data. One clear trend that can be

drawn from each property is that more electricity is required during winter months, when compared

to summer months. Two properties have been highlighted for further discussion as they exhibit

trends that differ significantly from the four other properties.

Figure 2.3-1: Annual usage trends for all six case studies supplied by EG.

The educational property displays the most significant monthly variation; the period of lowest usage

can be seen from July through August. Moreover, December observes a drop off in usage

(uncharacteristic of the typical winter usage levels when compared to the other properties). These

features align well with UK school closure periods; where the property will be unoccupied for long

periods of time during the given months. Peak usage can be seen in January when the property

reopens and heating requirements are at their highest. One point of note is that usage during

September is higher than March on average, despite official UK ‘summer time’ continuing until 22nd

September. This can most likely be attributed to electricity requirements associated with property

start up after the long summer holiday shut down, whereas the school would have been in full

operation for two months prior to March. The beginning of the Easter half term towards the end of

March may also play a major role in this incongruity.

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

11.0%

12.0%

13.0%

Pro

po

rtio

n o

f To

tal A

nn

ual

Usa

ge

Commercial Educational

Industrial Light Commercial

Residential Social


70

The residential property displays the most significant total difference between winter and summer

electricity usage. On average, each winter month accounts for 11.9% of the total annual usage,

compared to 5.0% during each summer month. This is largely due to increased domestic heating

requirements and the number of hours occupants spend indoors from December to February. The

most dramatic variation between subsequent months is April to May (falling from 10.0% to 6.8%

respectively). Even though May is officially classified as a spring month; solar irradiance is at its

highest during May when compared to any other month (see Figure 1.5-3). However, the increase in

solar irradiance from April to May alone is not enough to account for the significance of this change.

As mentioned previously, the residential case study is used as student accommodation, the 15/16

term dates for the university it serves run from 21st September to 20th May. The accommodation

contract however runs until 30th June. The departure of the occupants from the 20th May onwards

most likely accounts for the significant drop off in electricity usage from May until September.

All the remaining properties followed a reasonably similar trend with each summer month requiring

an average of 6.9% of the total annual usage, compared to 10.0% for the winter months. The

seasonal distribution of electricity usage for all properties sampled can be seen in Figure 2.3-2.

Winter usage is 12% larger than summer usage; autumn and spring demand is identical. If possible,

this figure suggests that all case studies would benefit from seasonal usage behaviour changes.

Figure 2.3-2: Average seasonal usage as a proportion of total annual usage for all properties in the sample analysed from half-hourly data.

Winter 31%

Spring 25%

Summer 19%

Autumn 25%

Proportion of Total Annual Usage


71

2.3.1 Weekly Usage Trends

Figure 2.3-3 displays the proportion of electricity used during each day of the week for the six case

study properties. This graph was created from the analysis of the EG supplied half hourly data and

displayed on the same axis for comparative purposes. Every property followed a trend of low

weekend usage, barring the social and residential properties. These two properties have been

highlighted. The weekly trends for each property will be discussed in further detail in section 2.4.

Figure 2.3-3: Weekly usage trends for all six property types from data supplied by EG.

Figure 2.3-4: Average usage as a proportion of total weekly usage for all properties.

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%

Pro

po

rtio

n o

f To

tal W

ee

kly

Usa

ge



Residential Social


72

2.3.2 Daily Usage Trends

Figure 2.3-5 is the plot of the proportion of electricity used during each half hour of the day for the

six case studies supplied by EG. The trends have been displayed on the same axis for comparative

purposes. Usage typically peaked once a day between 9AM and 12:00PM. The social and

residential properties did not follow this trend however. As can be seen, these properties either

witnessed a peak usage at a different time of the day or multiple usage peaks. The daily trends for

each property will be discussed in further detail in section 2.4. Despite accounting for only half of a

total day, the average 7AM to 7PM (Daytime) accounted for 63% of total daily usage for the

analysed case studies (see Figure 2.3-6).

Figure 2.3-5: Daily usage trends for all six property types from data supplied by EG.

Figure 2.3-6: Average usage as a proportion of total daily usage for all properties.

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge



Residential Social


73

2.4 Case Study Results

2.4.1 Commercial Property Usage Trends

87,600 data points were provided for the commercial property, encompassing 1st January 2011 to

31st December 2015. The total yearly usage increased from 43,190kWh in 2011 to 47,300kWh in

2015 (or 9.5% of the 2011 total).

From Table 2.4-1 it can be calculated that 53.6% of all electricity use was consumed during

weekdays, 20.3% during the night, 12.7% during evening times, and 13.4% during weekend days

before PV installation. E/W or W/N/E/W would be the cheapest pricing structure before PV

installation costing ≈ £445/annum. After PV installation, the weekday and weekend day

consumptions fell to 39.4% and 6.5% respectively. Under these circumstances the W/N/E/W

structure remained the cheapest option by a margin of £20/annum. The daytime consumption

offset by PV installation equals 21.1% of total daily consumption.

Table 2.4-1: Tariff cost breakdown for commercial property.

Structure

PV

Installe

d?**

Units Consumed (kWh)

Total Annual

Bill*

All

Day

Nigh

t

We

ekd

ays

Even

ing

We

eken

d

Unrestricted No 3800 - - - - - £470

Yes 1900 - - - - - £290

Day/Night No - 3030 770 - - - £460

Yes - 1340 570 - - - £280

Evening/Weekend No - - - 2810 480 510 £440

Yes - - - 1440 460 120 £270

Weekday/Night/Evening/Weekend No - - 770 2040 480 510 £430

Yes - - 570 750 460 120 £250 *Calculated from values given in Table 1.4-2 for property with average annual of 3,800kWh. **Calculated from estimated net meter values after installation of 4kWp PV system.

Figure 2.4-1 illustrates the significant drop off in usage for weekend days for the commercial

property. On average, Saturday and Sunday consume 6% of the total weekly consumption less than

any other weekday. The property is usually unoccupied during weekends, explaining this variation.

The standard error throughout the week is much smaller than the other case studies because of the

sheer volume of data supplied by EG for this case study. Despite this, standard error is slightly larger

for Monday when compared against the rest of the week. This is most likely because of national

holidays or bank holidays the fall on Mondays more frequently than any other day of the week.

These shut down periods reduce average demand, subsequently increasing standard deviation.


74

Figure 2.4-1: Average weekly usage trend for commercial property.

Figure 2.4-2 reflects the typical working hours for the commercial property reasonably well. As the

property allows staff to work flexitime hours, peak usage should, and does, range from 7AM to 6PM.

The standard error is also largest between the hours of 7AM-6PM. This is most likely down to

weekend shut downs, decreasing the average consumption during working hours and, consequently

increasing standard error. Continuously activated appliances (such as refrigerators) most likely

account for the evening and night consumption.

Figure 2.4-2: Average daily usage trend for commercial property.

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%P

rop

ort

ion

of

Tota

l We

ekl

y U

sage

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge

Time of Day


75

As the vast majority of consumption takes place during daylight hours already, it would be difficult to

reduce the export ratio from 5.5%. If there are any non-essential applications running outside of

daylight hours EG would encourage their usage during daylight hours (e.g. turning appliances in

standby mode off at the wall).

2.4.2 Educational Property Usage Trends

17,520 data points were provided by EG for the educational property, ranging from June 2013 to

May 2014. 57.1% of total consumption takes place during weekdays, 19.0% at night, 11.2% in the

evening, and 12.7% during weekend days before PV installation (as can be seen in Table 2.4-2).

Following PV installation, the total weekday and weekend day consumption levels falls to 46.4% and

6.3% respectively. Before and after PV installation the W/N/E/W pricing structure is the cheapest

option for the educational property; £10 less than alternate pricing structure in both instances at

£430 and £280 respectively. The daytime consumption offset after PV installation equals 20.1% of

total daily consumption.

Table 2.4-2: Tariff cost breakdown for educational property.

Structure

PV

Installe

d?**


Total Annual

Bill*

All

Day

Nigh

t

We

ekd

ays

Even

ing

We

eken

d

Unrestricted No 3800 - - - - - £470 Yes 2050 - - - - - £300

Day/Night No - 3080 720 - - - £460 Yes - 1490 560 - - - £290

Evening/Weekend No - - - 2890 430 480 £440 Yes - - - 1640 410 130 £290

Weekday/Night/Evening/Weekend No - - 720 2170 430 480 £430 Yes - - 560 950 410 130 £280

*Calculated from values given in Table 1.4-2 for property with average annual of 3,800kWh. **Calculated from net meter values after installation of 4kWp PV system.

Figure 2.4-3 exhibits a drop off in consumption for both days of the weekend. All other weekdays

have approximately equal electricity consumption levels (8% of total weekly usage higher than

typical weekend demand on average). The standard error for weekdays is much larger than

weekends because of long shut down periods throughout the year. During the Christmas and

summer holidays, the weekday consumption will fall to levels similar to what is seen during

weekends. This leads to higher weekday variation in consumption, increasing standard deviation

and error. Fewer data points were provided for this property, when compared to the commercial

property; therefore, the standard error is much larger than what is seen in Figure 2.4-1.


76

Figure 2.4-3: Average weekly usage trend for educational property.

Figure 2.4-4 highlights two periods of increasing consumption. Demand initially increases from 6AM

to 7AM, before increasing again from 8AM to a peak demand at 10:30AM. The first increase in

demand most likely marks the arrival of teachers and other staff members, with the second increase

attributed to the students arrival. The demand decreases gradually from its peak point until 7PM.

The departure of the students can clearly be seen immediately following 3PM. The evening and

night time usage is fairly constant. Once again, the large standard error during school opening hours

(seen in Figure 2.4-4) can be attributed to the frequent shutdown periods throughout the year.

Figure 2.4-4: Average daily usage trend for educational property.

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%

24.0%P

rop

ort

ion

of

Tota

l We

ekl

y U

sage

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge

Time of Day


77

As the vast majority of consumption takes place during daylight hours, it would be difficult to reduce

the export ratio from the identified 11.6%. If there are any non-essential applications running

outside of daylight hours EG would encourage their usage during daylight hours. It may be worth

considering some form of power storage system for generation during summer months. This will

increase the use of, otherwise exported, energy during periods of peak demand.

2.4.1 Industrial Property Usage Trends

Approximately 36,500 data points were supplied by EG for the industrial property, dating from

August 2012 to July 2013 and May 2014 to May 2015. A 19.9% increase in annual consumption was

seen from 2012 to 2015.

Table 2.4-3 demonstrates that 43.1% of the total electricity was consumed during weekdays, 25.9%

during the night, 19.4% during evening hours, and 11.7% during weekend daytime. After PV

installation, weekday consumption feel to 24.4% and weekend day consumption fell to 4.8% of the

daily total. Both before and after PV installation, the W/N/E/W is the cheapest pricing structure.

After PV installation the property would to save £60 per annum by switching to this pricing structure

from any other option available. The offset daytime consumption equals 25.6% of total daily

consumption.

Table 2.4-3: Tariff cost breakdown for industrial property.

Structure

PV

Installe

d?**


Total Annual

Bill*

All

Day

Nigh

t

We

ekd

ays

Even

ing

We

eken

d


Day/Night No - 2820 990 - - - £450 Yes - 1330 780 - - - £290




Figure 2.4-5 displays peak consumption for the industrial property; occuring during all weekdays,

barring Monday. On average, weekend days consume 5.7% of the total weekly consumption less

than Thursday through Friday. This is because, unlike the manufacturing staff members, most office

staff members do not work weekend shifts. There is no initially apparent link between the


78

operational hours of the industrial property and the low consumption levels exhibited for Mondays.

The standard error in these results is relatively small when compared to the other case studies. This

is most likely because this property only officially stops manufacture for three days of the year.

Moreover, twice the amount of data was provided for this property than the educational, light

commercial, and residential properties; further reducing the potential standard error in the results.

Figure 2.4-5: Average weekly usage trend for industrial property.

Figure 2.4-6: Average daily usage trend for industrial property.

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%

Pro

po

rtio

n o

f To

tal W

ee

kly

Usa

ge

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge

Time of Day


79

Figure 2.4-6 illustrates slightly higher levels of consumption between 8AM and 2PM. The increase in

demand commences from 6AM. From the peak seen at 9AM, demand gradually falls back to a

minimum at midnight. The plant operates on two 12 hour shifts starting at 7AM and 7PM for the

manufacturing staff; the office staff members work typical 9-5 hours. A combination of these

working hours most likely accounts for the gradual increase in consumption from 5AM to 9AM; the

shift workers will depart and arrive around 6AM, with the office staff progressively arriving in the

hours leading up to 9AM. The shift pattern would suggest a slight increase in demand around 6PM

as the night time shift arrives and day shift departs. This suggests that changes in demand are more

dependent upon the actions of the office staff. Knowing this helps focus any electricity usage

behavioural change initiatives towards the office rather than the manufacturing staff members.

Because 24 hour operation is in place, the standard error is roughly equivalent for all times in the

day. This error is most likely down to the highly variable energy intensity of the processes required

for production. It may be worth collecting a few years’ worth of additional data to see if the size of

the standard error is down to the relatively small collection period.

The export ratio is surprisingly low for the industrial property. Despite the significantly more evenly

spread daily consumption levels, the annual usage trend was also found to be much more consistent

from month to month. This means that during high summer generation periods, the electricity

demand is consistently high enough to only warrant the low 5.5% export ratio. In order to reduce

this further, the property may want to push initiatives that ensure all office appliances are switched

off overnight. They may also benefit from shifting heating requirements to daylight hours rather

than night time periods; this would also require the installation of heat storage methods or better

insulation.

2.4.2 Light Commercial Property Usage Trends

Over 17,500 data points were provided by EG for the light commercial property, ranging in date from

April 2009 to March 2010. This limited amount of outdated data means that any conclusions drawn

from this analysis will have relatively high error margins.

The proportional consumption levels are distributed as follows; 58.2% during weekdays, 17.0%

during the night, 12.9% during weekend days, and 12.0% during evening hours. E/W is the cheapest

pricing structure for this property before PV installation. After PV installation, the daytime

consumption levels would fall to 47.9% and 6.4%, for weekdays and weekends respectively, this led

the W/N/E/W pricing structure to become the cheapest option by an average margin of £70 (see

Table 2.4-4). The offset daytime consumption equals 16.8% of total daily consumption.


80

Table 2.4-4: Tariff cost breakdown for light commercial property.

Structure

PV

Installed

?**


Total Annual

Bill*

All

Day

Nigh

t

Weekd

ays

Evenin

g

Weeken

d


Day/Night No - 3160 650 - - - £460 Yes - 1570 510 - - - £300




Figure 2.4-7: Average weekly usage trend for light commercial property.

Figure 2.4-3 exhibits the most varied weekly usage of all properties analysed. Saturday sees the

lowest consumption levels, with an average of 8.1% of total weekly consumption for the year

analysed. This factor, combined with the relatively small standard error suggests that the property is

rarely occupied on Saturdays. Sundays show a slightly higher consumption level, with a much larger

standard error then any other day of the week. This suggests that the property is regularly occupied

on Sundays. Fridays consume ≈ 3% of the weekly consumption total less than the average of all

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%

24.0%

Pro

po

rtio

n o

f To

tal W

ee

kly

Usa

ge


81

remaining weekdays. This would suggest that fewer staff work Fridays than other weekdays, as the

standard error is not large enough to suggest regular Friday shut downs.

Figure 2.4-8 displays a clear 9-5 peak usage pattern. Peak demand occurs at 11AM, not until 3:30PM

does this demand start to decrease at a considerable rate. The evening and night time usage is

particularly stable and can be attributed to any devices or appliances active after working hours (e.g.

refrigerator). The standard error is substantially larger during working hours compared to evening

and night time hours. This is most likely due to weekend and holiday shut down periods, decreasing

the usage average, subsequently increasing error margins between 9AM and 5PM. The fact that

only one year’s worth of data is available may also lead to high standard error.

Figure 2.4-8: Average daily usage trend for light commercial property.

The vast majority of consumption takes place during daylight hours, complementing a low export

ratio. However, the seasonal variance (low summer demand) leads to a relatively high export ratio

of 11.8%. It would be difficult to reduce the export ratio from its current level from changes to daily

usage patterns. Instead, significant electricity requirement shifts would need to be seasonal (from

winter to summer) to reduce the calculated export ratio. Much like the educational property, a

financial assessment on the feasibility of a power storage or battery system may show benefits over

the current scenario.

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge

Time of Day


82

2.4.3 Residential Property Usage Trends

Over 17,500 data points were supplied by EG for the residential property, ranging in date from May

2010 to April 2011. More recent data will presumably be needed for a more accurate estimation of

current usage trends; however, the available data acts as a good starting point.

The electricity was consumed across a typical day as follows; 37.6% during weekdays, 24.7% during

evening hours, 23.1% during the night, and 14.6% during weekend days (all before PV installation).

Due to the relatively low daytime consumption, it stands to reason that W/N/E/W pricing structure

is the cheapest option for the residential property. After PV installation, the daytime consumption

distribution was estimated to fall to 22.8% and 8.7%, for weekday and weekend periods respectively.

The W/N/E/W pricing structure remains the cheapest option for the residential property

(£220/annum). At £100/annum cheaper than the next cheapest option, no other property sees such

significant benefit from changing their pricing structure. The daytime consumption offset by PV

installation equals 20.7% of total daily consumption.

Table 2.4-5: Tariff cost breakdown for residential property.

Structure

PV

Installe

d?**


Total Annual

Bill*

All

Day

Nigh

t

We

ekd

ays

Even

ing

We

eken

d


Day/Night No - 2920 880 - - - £450 Yes - 1670 730 - - - £320




Figure 2.4-9 exhibits the most consistent weekly consumption of any of the sampled properties.

Opposed to all the other case studies, there is no trend suggesting higher or lower consumption

levels during the weekend period. This suggests that the occupancy hours for every day of the week

are roughly equal for student lifestyles. The standard error margins for these results are relatively

large because of the long summer vacancy (this property is mainly used as student accommodation).

This lowers the average weekly usage significantly, subsequently increasing error margins. This is

also symptomatic of the relatively few data points supplied for this case study used to generate the

usage trends.


83

Figure 2.4-9: Average weekly usage trend for residential property.

Figure 2.4-10: Average daily usage trend for residential property.

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%

Pro

po

rtio

n o

f To

tal W

ee

kly

Usa

ge

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge

Time of Day


84

Consumption is distributed relatively evenly throughout all times of the day for the residential

property (see Figure 2.4-10). Usage peaks at 7:30PM, only 1.3% of the total daily consumption

higher than the period of lowest demand from 1:30AM to 4:30AM. Consumption gradually increases

from 5AM to 8AM, where it remains relatively stable until 4PM; at which time, a period of increasing

demand until 7:30PM is witnessed. The morning increase can be attributed to the occupants

progressively waking up and activating electrical appliances. Correspondingly, the afternoon into

evening increase is presumably caused by students returning from their studies and using appliances

such as laptops, phone chargers, electric ovens, televisions, and lighting. The standard error is large

relative to the average values (>90% of the average at every point). This is almost certainly a result

of the summer vacancy period and relatively small data set provided.

This property has the highest export ratio of all the case studies (21.2%). The daily trend highlights

the need for a higher proportion of daylight hour demand. If achieved, this would overcome the

largest factor of the high predicted export ratio. Where possible, occupants should use energy

intensive appliances (such as washing machines, clothes dryers, and space heaters) when the sun is

up. The landlord can invest in better insulation that would lessen heating requirements later in the

day, but there is a limit to the potential export ratio reduction. The summer shut down periods

stress the need for seasonal consumption shifts. Investment in energy storage methods such as the

Tesla Powerwall may be the only way to gain an export ratio of 0% for this property (Rodrigues,

Faria, Ivaki, Cafofo, Chen, & Morgado-Dias, 2016).

2.4.4 Social Property Usage Trends

Approximately 65,000 data points were provided by EG for the social property, dating from

September 2010 to October 2015. Total yearly consumption increased 4.8% over this time period

(from 7150kWh in 2011 to 7500kWh in 2015). Further information would be required in order to

develop forecasted usage increases over the lifetime of a solar PV generation; however this could

prove to be an interesting potential future study.

40.3% of all electricity was consumed during weekdays, 23.8% during the night, 19.3% during

weekend days, and 16.6% during evening hours. The W/N/E/W pricing structure was estimated to

be at least £30 cheaper than all other options before PV installation (£420/annum). Following PV

installation, the daytime consumption was estimated to fall to 25.5% and 13.6% of the total daily

consumption, for weekday and weekend days respectively. Once again the W/N/E/W option was

the cheapest option at £280/annum or £30 less than all other structures on average. The total

daytime consumption offset by PV installation equalled 20.7% of total daily consumption.


85

Table 2.4-6: Tariff cost breakdown for social property.

Structure

PV

Installed

?**


Total Annual

Bill*

All

Day

Nigh

t

Weekd

ays

Evenin

g

Weeken

d


Day/Night No - 2900 900 - - - £450 Yes - 1460 730 - - - £300




Figure 2.4-11 exhibits higher consumption levels on Saturdays than any other day of the week, 4.8%

of total weekly consumption higher than the Monday-Thursday average. Friday and Sunday also

display higher than average consumption levels. This was expected as the property is frequented by

club members on these three days more than any other. The standard error for these results is

relatively small in comparison to the other case studies, because of the relatively large amount of

data provided by EG.

Figure 2.4-11: Average weekly usage trend for social property.

0.0%

4.0%

8.0%

12.0%

16.0%

20.0%

24.0%

Pro

po

rtio

n o

f To

tal W

ee

kly

Usa

ge


86

Figure 2.4-12 shows widely distributed consumption throughout an average day for the social

property. Two separate peaks in usage can be seen at 12PM and 7PM. This can be attributed to the

catering times and closing times of the property. Consumption first increases from 12AM to 3AM

(possibly due to the initiation of time activated appliances). No further increase in consumption is

then seen until 7AM (when staff members start to arrive and set up for the customers). The catering

service hours are 12PM-4PM Sunday through Wednesday, and 12PM-8PM Thursday through

Saturday. Averaging these service hours may cause the slight drop off in usage until 5PM, and

second peak at 7PM.

The clubhouse bar remains open until 11PM most nights; however its electricity requirements are

not significant enough to increase the evening hours’ consumption levels (as seen in Figure 2.4-12).

Despite the abundance of data provided by EG, the standard error for these results is relatively

large. This is most likely caused by the fluctuating seasonal opening hours. However, this does not

account for the large standard error seen before 9AM; as the demand during this period of time

should be unaffected by occupancy. There may well be time activated appliances that are also

seasonally dependant (e.g. sprinklers in the summer, heaters during winter) at play.

Figure 2.4-12: Average daily usage trend for social property.

The 12.0% export ratio predicted for this property will be difficult to reduce as this property’s

consumption is customer behaviour dependant. The energy intensive catering services respond to

customer orders, only the time activated appliances can be feasibly altered to limit night time usage.

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

00:00 04:00 08:00 12:00 16:00 20:00

Pro

po

rtio

n o

f To

tal D

aily

Usa

ge

Time of Day


87

The club could offer deals convincing customers to order meals during daylight hours; however this

may affect customer turnout. A full cost/benefit analysis should be completed for proceeding with

any customer facing changes.

2.4.5 Case Study Summary

A summary of the results and discussion for the six case studies can be seen in Table 2.4-7;

Table 2.4-7: Summary of case study key results discussed in section 2.4.

Property Type Average Peak Usage Cheapest Pricing Structure Offset Demand

during ‘day’ Month Day Time Pre-PV Post-PV

Commercial February Monday 9AM W/N/E/W W/N/E/W 21.1%

Educational January Wednesday 10:30AM W/N/E/W W/N/E/W 20.1%

Industrial February Wednesday 9AM W/N/E/W W/N/E/W 25.6%

Light Commercial January Tuesday 11AM E/W W/N/E/W 16.8%

Residential February Monday 7:30PM W/N/E/W W/N/E/W 20.7%

Social January Saturday 12PM W/N/E/W W/N/E/W 20.7%


88

3.0 Conclusions

The total amount that FiT payments contribute to the lifetime profits of PV systems will fall from

56.3% to 30.9% upon the introduction of the proposed 4.39p/kWh tariff in February 2016.

Correspondingly, the total contribution to lifetime profits that self-consumption savings make will

increase from 33.4% to 52.8%. The payback period for the modelled system will increase by ≈ 32.5%

from 9.6 years to 12.7 years, assuming identical cost and efficiency to 2015 values, under the new

legislation.

On average, each class of solar assessment tool required similar specification levels (7 inputs for

primary solar solution providers versus 6.7 inputs for the remaining sample). The required number

of inputs ranged from 10 to 4 for the sample, whereas EG’s final tool allows the definition of 19 input

variables. Of all results displayed by commercially available solar assessment tools, 75% were

financial in nature; the remaining 25% of results were environmental in nature. On average, the

primary solar solution providers displayed 5 results, versus 6 results for the secondary solar solution

providers and not-for-profit organisations. The number of results displayed ranged from 9 to 3,

whereas EG’s tool displays 20 individually discernible results. The ratio between financial and

environmentally based results was 1:1 for EG’s tool. In conclusion, EG’s tool requires the highest

level of specification (increasing prediction precision) and reports the most information to the user

(reducing ambiguity through transparency).

The primary solar solution providers predict a lifetime profit 145% higher than the remaining

sample. Notably, the not-for-profit organisations predict a lifetime profit 3.6% higher than

secondary solar solution providers. EG’s predicted profit was 7.7% less than the sample average

when identical assumptions are applied; the results were also found to fall well within the standard

error range. The average profit for the property types was 17.5% higher than the £7,700 sample

average. Estimated self-consumption savings varied significantly more than estimated FiT and

export payments (standard deviation equalled 44% of sample mean versus 3% and 16%

respectively). When applying a 50% self-consumption assumption, EG’s self-consumption saving

prediction was 23% higher than the alternative calculators.

Very little variance between alternative calculators was seen for ‘First year Generation’ predictions

(standard error of 49.7kWh for average 3170kWh generation). However, EG’s generation

predictions were found to consistently fall short of the acceptable minimum (calculated from the

standard error). When using identical assumptions, EG’s generation predictions were 17.9% less

than the alternative calculator average.


89

All property types were found to have export ratios well below the 50% assumption employed by

alternative calculators for the assessed system. Ranging from 5.5% to 21.2% of total generation, the

export ratio was highest for the residential property. If able to reduce the export ratio to 0%, the

residential property could potentially increase lifetime profits by £820; the most of any property.

For all properties in the sample the total exported electricity during summer months equalled 53.8%

of total annual export. Spring accounted for the next highest proportion of annual export with

33.3%, followed by 12.8% during autumn months. No energy was exported during winter months,

barring the light commercial property; however, this still only accounted for 0.3% of the total annual

export.

The annual electricity usage trends all followed a consistent pattern of higher usage during the

winter, and lower usage during the summer. The residential property displays the highest seasonal

variance with 170% more electricity used in February compared to August. The educational

property displayed the most monthly variance in monthly usage. For the total weekly usage, 77%

was used during weekdays for the six case studies. Of the total daily usage, 63.4% was used from

7AM to 7PM for the assessed properties, with 16% used from 7PM to midnight, and 20.7% used

from midnight to 7AM.

The weekday/night/evening/weekend was the cheapest pricing structure for all the properties after

PV installation. All properties could benefit from changing usage behaviours to lower net export

levels. However, it would be difficult for the residential property to entirely eliminate all periods of

net export, because of the summer vacancies.

For complete case study assessment results see section 6.5. The result print-outs outline all the

factors defined for the system being assessed, as well as the customer facing results.


90

4.0 Future Developments

Going forward, there are several tariffing additions that would improve the usability of EG’s

assessment tool. EG has access to the MPAN number of their clients. The MPAN number contains

all the information needed to fully define the electricity pricing structure of a property. Writing a

programme that can translate any given MPAN will give more accurate self-consumption saving

predictions. The regular pricing structure updates EG receive from each DNO also means the current

values used in calculations will become obsolete in the near future; the need for constant

maintenance make it preferable to have some mechanism that monitors and updates the tool

automatically from DNO correspondence.

Currently, the tool compares four major pricing structures to identify and recommend the cheapest

option for the property. Mentioned previously, there are several hundred pricing structure options

available to the consumer. Creating a system that collects the latest MTC data from each DNO and

recommends the cheapest option to the user would be a great advancement on what is currently

offered. The potential MPAN programme described above will allow the exclusion of any pricing

structure ineligible for the assessed property.

Data from several examples of each property type should also be analysed to develop ‘typical usage

trends’. Currently, the tool has the data for six unique case studies; however, using these examples

to make predictions for alternative properties would be inaccurate. In order to reduce the standard

error seen in the usage results, several years worth of data for each property is required. Users may

feel that the property being assessed does not fall into any of the categories defined by the tool and

may prefer a different option instead. Therefore, creating a ‘general’ property type by averaging all

the data supplied by EG for this project may be of value.

Currently, all property usage analysis is completed in a separate programme to the solar assessment

tool. It might be useful to combine these two programmes, allowing the user to select ‘bespoke’ as

the property type. Simplifying the analysis process would then allow the user to input their own

historical usage data, rather than rely upon EG’s auditor for a bespoke quote. However, this may be

a service that EG wishes to remain in-house.

In terms of the solar assessment calculations, it would be conducive if the tool found an alternative

to the MCS specified irradiance data. Over time, with changing climate and weather patterns the

data supplied by the MCS will become outdated. Instead, mathematic relationships between

irradiance and the system longitude, latitude, inclination, and orientation could be used to calculate

the expected kWh/kWp values; rather than looking up these values from tables as the tool currently


91

does. Moreover, this would allow users to assess solar PV installations outside the UK. One

potential source of this data would be NASA’s surface meteorology and solar energy tables derived

from 22 years of monthly averaged readings taken from 200+ satellites.

EG hope the tool can accurately assess the potential of PV system for a wide range of sizes and

specifications. For this reason, it will be important to investigate the precise effects of tracking

systems on expected generation; once this data has been gathered and interpreted it can become a

supplementary input for the user. The wide range of systems commercially available also makes it

important to research kWh/kWp values for cells types other than silicon PV. Given enough time, EG

hope to create an assessment tool for solar thermal installations as well.

Finding such a considerable difference in results between the supposedly ‘impartial’ alternative

calculators highlights the importance of transparency during the assessment process. EG should

initiate conversations with the developers of the calculators selected for comparison, disclosing the

results of this study. This may invoke a dialogue between each company, encouraging cross party

co-operation. The majority of the alternative calculators sampled are yet to update to the most

recent FiT and export tariffs; if anything, this study should convince other developers to address this

oversight.

EG’s ambition to create the most accurate solar assessment tool commercially available requires

rigorous testing to determine whether this goal has been achieved. Access to the half hourly data of

a property before and after PV installation will allow this testing. From the analysis of said

property’s usage behaviours and the entry of the system specifications, EG will be able to see

whether their net metering predictions mimic the meter readings for the years proceeding PV

installation.


92

5.0 References

Abdalla, I. A. (2013) Integrated PV and Multilevel Converter System for Maximum Power Generation

under Partial Shading Conditions. Leeds, The University of Leeds.

Alsema, E. A. et al. (2006) Environmental impacts of PV electricity generation - a critical comparison

of energy supply options ECN. 21st European Photovoltaic Solar Energy Conference and Exhibition.

Dresden.

Alter, L. (2015) Bike lane down centre of Korean highway is covered with solar panels [online].

Available from: www.treehugger.com/bikes/bike-lane-down-center-korean-highway-covered-solar-

panels.html [Accessed 26th January 2016].

Alves, L., & Boling, N. (2010) High Efficiency Solar Coatings. Santa Rosa, Novus Media Today.

ASTM (2012) Standard Tables for Reference Solar Spectral Irradiance at Air Mass 1.5: Direct Normal

and Hemispherical for a 37 Tilted Surface. Denver, American Society for Testing and Materials.

Auger, P. (1923) The compound photoelectric effect (Sur les rayons β secondaires produits dans un

gaz par des rayons X). C.R.A.S, 177(1), pp. 169-171.

Barrows, A. T., Pearson, A. J., Kwak, C. K., Dunbar, A. D., Buckley, A. R., & Lidzey, D. G. (2014) Efficient

planar heterojunction mixed-halide perovskite solar cells deposited via spray-deposition. Energy &

Environmental Science, 7(1), pp. 2944-2950.

Blakers, A., & Weber, K. (2000) The Energy Intensity of Photovoltaic Systems. Canberra, Centre for

Sustainable Energy Systems.

Campoccia, A. et al. (2009) Comparative analysis of different supporting measures for the production

of electrical energy by solar PV and Wind systems: Four representative European cases. Solar Energy,

83(1), pp. 287-297.

CAT (2016) What is the energy and carbon payback time for PV panels in the UK? [online]. Available

from: info.cat.org.uk/questions/pv/what-energy-and-carbon-payback-time-pv-panels-uk [Accessed

15th April 2016].

CAT (2016) Mission Statement [online]. Available from: content.cat.org.uk/index.php/mission-

statement [Accessed 5th January 2016].

Collavini, S. et al. (2015) Understanding the Outstanding Power Conversion Efficiency of Perovskite-

Based Solar Cells. Angewandte Chemie International Edition, 54(34), pp. 9757-9759.


93

DECC (2015). Solar Photovoltaics Deployment in the UK. London, Department for Energy & Climate

Change.

EG-Audit Ltd. (2016) Pricing Matrices January 2016. Chester, EG-Audit Ltd.

EG-Audit Ltd. (2015) Industrial Pricing Structure Data Oct 16. Chester, EG-Audit Ltd.

Electrical Contractors Association (2012) Guide to the Installation of Photovoltaic Systems. London,

Microgeneration Certification Scheme.

Energy Initiative Massachusetts Institute of Technology (2015) Study on the Future of Solar Energy.

Massachusetts, MIT.

Energy Linx (2016) MPAN (Meter Point Administration Number) [online]. Available from:

www.energylinx.co.uk/mpan.htm [Accessed 17th February 2016].

Energy Saving Trust (2016) What we do [online]. Available from:

www.energysavingtrust.org.uk/what-we-do [Accessed 5th January 2016].

EPA (2012) Inventory of US Greenhouse Gas Emissions and Sinks: 1990-2010. Washington DC,

Environmental Protection Agency.

EPIA (2014) Global Market Outlook for Photovoltaics 2015-2019. Brussels, SPE.

Equilibre Personnel (2016) Solar Water Heater [online]. Available from: equilibre-personnel.com/tag-

solar-water-heater.html?adb=1 [Accessed 11th January 2016].

EPIA (2011) Global Market Outlook for photovoltaics until 2015. Brussels, European Photovoltaic

Industry Association.

Feed-in Tariffs (2016) Export tariffs [online]. Available from:

www.fitariffs.co.uk/FITs/principles/export/ [Accessed 4th February 2016].

Feed-in Tariffs (2016) Tariffs payable per kWh of electricity produced [online]. Available from:

www.fitariffs.co.uk/eligible/levels/ [Accessed 4th February 2016].

Fewins, C. (2012) Getting the Right Pitch [online]. Available from: www.homebuilding.co.uk/getting-

the-right-pitch/ [Accessed 11th March 2016].

FHWA (2013) Highway Statistics 2011: Table VM-1 [online]. Available from:

www.fhwa.dot.gov/policyinformation/statistics/2011/pdf/vm1.pdf [Accessed 23rd March 2016].


94

Fraunhofer Institute for Solar Energy Systems (2015) Photovoltaics Report. Freiburg, Fraunhofer ISE.

Fthenakis, V. et al. (2008) Emissions from Photovoltaic Life Cycles. Environmental Science &

Technology, 42(1), pp. 2168-2174.

Gardner, M. (2015) Average annual domestic standard electricity bills in 2015 for UK regions with

average unit costs. London, DECC.

Gauke, D. (2015) Budget 2015. London, HM Treasury.

Green, M. A. (1995) Optical Properties of Intrinsic Silicon at 300 K. Progress in Photovoltaics:

Research and Applications, 3(1), pp. 189-192.

Green, M. A. (2009) The path to 25% silicon solar cell efficiency: History of silicon cell evolution.

Progress in Photovoltaics: Research and Applications, 17(3), pp. 183-189.

Guerrini, F. (2016) France Wants To Install 1,000 Km Of Solar Roadways Over The Next Five Years

[online]. Available from: www.forbes.com/sites/federicoguerrini/2016/02/07/france-wants-to-

install-1000-km-of-solar-roadways-over-the-next-five-years/#4eec0287857e [Accessed 15th February

2016].

Hasuike, H. et al. (2006) Study on design of molten salt solar receivers for beam-down solar

concentrator. Solar Energy, 80(1), pp. 1255-1262.

Helwa, N. H. et al. (2000) Maximum Collectable Solar Energy by Different Solar Tracking Systems.

Energy Sources, 22(1), pp. 23-34.

Hoffman, W. (2006) PV solar electricity industry: Market growth and perspective. Solar Energy

Materials and Solar Cells, 90(19), pp. 3285-3311.

Honsberg, C., & Bowden, S. (2015) Welcome to PVCDROM [online]. Available from:

www.pveducation.org/pvcdrom [Accessed 2nd November 2015].

IEA (2014) Technology Roadmap: Solar Photovoltaic Energy. Paris, International Energy Agency.

Javid, S. (2014) Budget 2014. London, HM Treasury.

Karatepe, E. et al. (2007) Development of a suitable model for characterising photovoltaic arrays

with shaded solar cells. Solar Energy, 81(8), pp. 977-992.

Kasten, F., & Young, A. T. (1989) Revised optical air mass tables and approximate formula. Applied

Optics, 28(1), pp. 4735-4738.


95

Keoleian, G. A., & Lewis, G. M. (1997) Application of Life-cycle Energy Analysis to Photovoltaic

Module Design. Progress in Photovoltaics: Research and Applications, 5(1), pp. 287-300.

Kittel, C. (1986) Introduction to Solid State Physics (6th edition). New York, John Wiley & Sons.

Kojima, A. et al. (2009) Organometal Halide Perovskites as Visible-Light Sensitizers for Photovoltaic

Cells. Journal of the American Chemical Society, 131(17), pp. 6050-6051.

Lelouxa, J. et al. (2012) Review of the performance of residential PV systems in France. Renewable

and Sustainable Energy Reviews, 16(1), pp. 1369-1376.

Lenzen, M. (2008) Life Cycle and greenhouse gas emissions of nuclear energy. Energy conversion and

management, 39(1), pp. 2178-2199.

Lonsdale, S. (2015). Eco-living: Beware the 'solar-panel cowboys' [online]. Available from:

www.telegraph.co.uk/finance/property/9724311/Eco-living-Beware-the-solar-panel-cowboys.html

[Accessed 5th November 2015].

Lucas, H. (2014) Quarterly Energy Prices- December 2014: Household Energy Prices. London,

Department of Energy & Climate Change.

Lucas, H. (2015) Average annual domestic electricity bills for selected towns and cities in the UK and

average unit costs (QEP 2.2.3). London, Department of Energy & Climate Change.

Lui, Y. et al. (2010) Research on an adaptive solar photovoltaic array using shading degree model-

based reconfiguration algorithm. Control and Decision Conference. Xuzhou.

Meinel, A. B., & Meinel, M. P. (1976) Applied Solar Energy: An Introduction (1st edition). Boston,

Addison Wesley Publishing Co.

Met Office (2014) Solar Energy [online]. Available from: www.metoffice.gov.uk/renewables/solar

[Accessed 6th January 2016].

Midsummer Energy. (2016). Optimum angle for solar panels [online]. Available from:

midsummerenergy.co.uk/solar-panel-information/Utilities/SolarPanelsOptimumAngle [Accessed 11th

March 2016].

Migo, E. (2014) Monthly deployment of all solar photovoltaic capacity in the United Kingdom.

London, Office of National Statistics.


96

Mufson, S. (2013) Prices Flat in Polysilicon Market [online]. Available from:

www.washingtonpost.com/business/economy/prices-flat-in-polysilicon-

market/2013/07/23/914479d0-f3e4-11e2-9434-60440856fadf_graphic.html [Accessed 13th January

2016].

Muhammad-Sukki, F. et al. (2013) Revised feed-in tariff for solar photovoltaic in the United Kingdom:

A cloudy future ahead? Energy Policy, 52, pp. 832-838.

Muneer, T. (2004) Solar Radiation and Daylight Models (2nd edition). Oxford, Elsevier Butterworth-

Heinemann.

NASA (2016) NASA Surface meteorology and Solar Energy - Available Tables [online]. Available from:

eosweb.larc.nasa.gov/cgi-

bin/sse/grid.cgi?&num=177142&lat=51.4&submit=Submit&hgt=100&veg=17&sitelev=&email=skip

@larc.nasa.gov&p=grid_id&p=clr_sky&p=day_cld&step=2&lon=-3.2 [Accessed 21st March 2016].

Noel, N. K. et al. (2014) Lead-free organic–inorganic tin halide perovskites for photovoltaic

applications. Energy & Environmental Science, 7(9), pp. 3061.

NREL (2015) Best Research- Cell Efficiencies [online]. Available from:

www.nrel.gov/ncpv/images/efficiency_chart.jpg [Accessed 5th January 2016].

NREL (2015) Planta Solar 20 [online]. Available from:

www.nrel.gov/csp/solarpaces/project_detail.cfm/projectID=39 [Accessed 29th January 2016].

OFGEM (2016) The GB electricity distribution network [online]. Available from:

www.ofgem.gov.uk/electricity/distribution-networks/gb-electricity-distribution-network [Accessed

4th February 2016].

Patel, B. C. (2010) Photovoltaic Cells: Converting Photons to Electricity [online]. Available from:

eng3060.pbworks.com/w/page/18919015/Photovoltaic%20Cells%3A%20Converting%20Photons%2

0to%20Electricity [Accessed 13th January 2016].

Peng, J., Lu, L., & Yang, H. (2013) Review on life cycle assessment of energy payback and greenhouse

gas emission of solar photovoltaic systems. Renewable and Sustainable Energy Reviews , 19(1), pp.

255-274.

Perez, R., Ineichen, P., & Seals, R. (1990) Modelling Daylight Availability and Irradiance Components

from Direct and Global Irradiance. Solar Energy, 44(5), pp. 271-289.


97

Plunkett, C. (2011) National Renewable Energy Lab in Golden to cut 100-150 jobs through buyouts

[online]. Available from: www.denverpost.com/news/ci_19034058 [Accessed 5th January 2016].

PV & SOLAR EDT (2016) How to calculate the annual solar energy output of a photovoltaic system

[online]. Available from: photovoltaic-software.com/PV-solar-energy-calculation.php [Accessed 23rd

January 2016].

PVCDROM (2016) The Sun's Position [online]. Available from:

www.pveducation.org/pvcdrom/properties-of-sunlight/suns-position [Accessed 28th February 2016].

Ramabadran, R., & Mathur, B. (2009) Effect of Shading on Series and Parallel Connected Solar PV

Modules . Modern Applied Science, 3(10), pp. 32-41.

Randall, C. (2011) Housing- Social Trends. London, Office for National Statistics.

Sawin, J. (2014) Renewables 2014: Global Status Report. Paris, REN21.

Renewable Energy Consumer Code (2015) Feed in Tariff Scheme [online]. Available from:

www.recc.org.uk/pdf/feed-in-tariff-scheme-guidance-for-consumers.pdf [Accessed January 2016].

Renewable Energy World (2006) SolarWorld Acquires Shell's Solar Business [online]. Available from:

www.renewableenergyworld.com/articles/2006/02/solarworld-acquires-shells-solar-business-

42840.html [Accessed 5th January 2016].

Rhode, R. (2016) Solar Spectrum [online]. Available from:

www.globalwarmingart.com/wiki/File:Solar_Spectrum_png [Accessed 4th February 2016].

Rodrigues, S. et al. (2016) Tesla Powerwall: Analysis of its use in Portugal and United States.

International Journal of Power and Energy Systems, 36(1), pp. 7-12.

Samarakoon, K., Ekanayake, J., & Jenkins, N. (2011) Investigation of Domestic Load Control to

Provide Primary Frequency Response Using Smart Meters. IEEE Transactions on Smart Grid, 3(1), pp.

282-292.

Sedona Solar Technology (2016) Array Type [online]. Available from: sedonasolartechnology.com

[Accessed 25th March 2016].

SEMI (2014) International Technology Roadmap for Photovoltaic 2014 Results. Frankfurt, ITRPV.

Shah, V., Booream-Phelps, J., & Min, s. (2014) 2014 Outlook: Let the Second Gold Rush Begin.

Frankfurt, Deutsche Bank AG.


98

Shockley, W., & Queisser, H. J. (1961) Detailed Balance Limit of Efficiency of p-n Junction Solar Cells.

Journal of Applied Physics, 32(1), pp. 510-519.

Shockley, W., & Read, W. T. (1952) Statistics of the Recombination of Holes and Electrons. Physical

Review, 87(1), pp. 835.

SkyFuel (2011) Next Generation Solar Parabolic Trough Technology [online]. Available from:

www.skyfuel.com/downloads/brochure/SkyTroughBrochure.pdf [Accessed 11th January 2016].

Solar Century Ltd (2016) About SolarCentury [online]. Available from:

www.solarcentury.com/uk/about-solarcentury/ [Accessed 5th January 2016].

Solar Guide (2015) Terms and Conditions of Business- Leads [online]. Available from:

www.solarguide.co.uk/legal [Accessed 5th January 2016].

Solar Insolation (2012) Solar Insolation [online]. Available from: solarinsolation.org/ [Accessed 6th

January 2016].

Solar Panels UK (2016) Solar PV Calculator [online]. Available from: www.solarpanelsuk.co.uk/solar-

pv-calculator.php [Accessed 15th January 2016].

Solar World UK (2016) Welcome to our UK solar power calculator! [online]. Available from:

www.solarworld-uk.co.uk/service/solar-pv-calculator/ [Accessed 15th January 2016].

Sust-it (2016) Boiling a Kettle Cost Calculator [online]. Available from: www.sust-it.net/boiling-kettle-

cost-calculator.php [Accessed 10th February 2016].

Swanson, R. (2006) A vision for crystalline silicon photovoltaics. Progress in Photovoltaics: Research

and Applications, 14(1), pp. 443-453.

Taylor, J., Leloux, J., Hall, L. M., Everard, A. M., Briggs, J., & Buckley, A. (2015) Performance of

Distributed PV in the UK: a Statistical Analysis of over 7000 Systems. Hamburg, Sheffield Solar.

The ECOEXPERTS (2016) Solar PV Calculator 2015 [online]. Available from:

www.theecoexperts.co.uk/solar-pv-calculator?sid=160077233 [Accessed 5th January 2016].

The GreenAge (2016) The cost of a solar PV system [online]. Available from:

www.thegreenage.co.uk/tech/the-cost-of-a-solar-pv-system/ [Accessed 2nd March 2016].


99

Ubisse, A., & Sebitosi, A. (2009) A new topology to mitigate the effect of shading for small

photovoltaic installations in rural sub-Saharan Africa. Energy Conversion and Management , 50(1),

pp. 1797–1801.

UK Meteorological Office (1972) Part III. Europe and the Azores. In Tables of Temperature, Relative

Humidity, Precipitation and Sunshine for the World. London, HMSO.

UNFCCC (2015) Adoption of the Paris agreement—Proposal by the President—Draft decision.

Available from: unfccc.int/resource/docs/2015/cop21/eng/l09.pdf [Accessed 20th February 2016].

Vatansever, D., Siores, E., & Shah, T. (2012) Alternative Resources for Renewable Energy:

Piezoelectric and Photovoltaic Smart Structures. In: B. R. Singh (Ed.) Global Warming – Impacts and

Future Perspective. Bolton, Vatansever. pp. 263-290.

Vignola, F., Mavromatakis, F., & Krumsick, J. (2011) Performance of PV Inverters. Oregon, Solar

Radiation Publishing Laboratory.

Vos, A. D. (1980) Detailed Balance limit of the efficiency of tandem solar cells. Journal of Applied

Physics, 13(5), pp. 839-846.

Werner, J., Kolodinski, S., & Queisser, H. (1994) Novel Optimization Principles and Efficiency Limits

for Semiconductor Solar Cells. Physical Review Letters, 72(24), pp. 3851-3854.

Working Group III, IPCC (2014) Mitigation of Climate Change, Annex II I:Technology- specific cost and

performance parameters. London, Intergovernmental Panel on Climate Change.

Zhu, G. et al. (2013) History, current state, and future of linear Fresnel concentrating solar collectors.

Solar Energy, 103(1), pp. 639-652.


100

6.0 Appendices

6.1 Solar PV Generation Appendices

Figure 6.1-1: Reflectivity of a polished silicon wafer for the majority of the wavelengths of the visible spectrum (Green M. A., 1995).

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

200 300 400 500 600 700 800 900 1000

Ref

lect

ivit

y

Wavelength (nm)


101

6.2 Life Cycle Appendices

Figure 6.2-1: Manufacturing process for silicon based PV modules (Peng, Lu, & Yang, 2013).


102

6.3 Raw Data for tool

Figure 6.3-1: Example of half hourly data supplied by EG. All values are given in kWh/half hour.


103

Table 6.3-1: Postcode zoning data (Electrical Contractors Association, 2012).

Postcode Zone Postcode Zone Postcode Zone Postcode Zone Postcode Zone

AB 16 HG 10 LL60 13 PO 3 SA66 13

AL 1 HP 1 LL61 13 PO18 2 SA67 13

B 6 HR 6 LL62 13 PO19 2 SA68 13

BA 5E HS 18 LL63 13 PO20 2 SA69 13

BB 7E HU 11 LL64 13 PO21 2 SA70 13

BD 11 HX 11 LL65 13 PO22 2 SA71 13

BD23 10 IG 12 LL66 13 PO23 3 SA72 13

BD24 10 IP 12 LL67 13 PR 7E SA73 13

BD25 11 IV 17 LL68 13 RG 1 SA74 5W

BH 3 IV30 16 LL69 13 RG21 3 SE 1

BL 7E IV31 16 LL70 13 RG22 3 SG 1

BN 2 IV32 16 LL71 13 RG23 3 SK 7E

BR 2 IV33 17 LL72 13 RG24 3 SK13 6

BS 5E IV34 17 LL73 13 RG25 3 SK14 7E

BT 21 IV36 16 LL74 13 RG26 3 SK17 6

CA 8E IV37 17 LL75 13 RG27 3 SK18 7E

CB 12 KA 14 LL76 13 RG28 3 SK22 6

CF 5W KT 1 LL77 13 RG29 3 SK23 6

CH 7E KW 17 LL78 13 RG30 1 SK24 7E

CH5 7W KW15 19 LL79 7W RH 1 SL 1

CH6 7W KW16 19 LN 11 RH10 2 SM 1

CH7 7W KW17 19 LS 11 RH11 2 SN 5E

CH8 7W KW18 17 LS24 10 RH12 2 SN7 1

CH9 7E KY 15 LS25 11 RH13 2 SN8 5E

CM 12 L 7E LU 1 RH14 2 SO 3

CM21 1 LA 7E M 7E RH15 2 SP 5E

CM22 1 LA7 8E ME 2 RH16 2 SP6 3

CM23 1 LA8 8E MK 1 RH17 2 SP7 3

CM24 12 LA9 8E ML 14 RH18 2 SP8 3

CO 12 LA10 8E N 1 RH19 2 SP9 3

CR 1 LA11 8E NE 9E RH20 2 SP10 3

CT 2 LA12 8E NG 11 RH21 1 SP11 3

CV 6 LA13 8E NN 6 RH77 2 SP12 5E

CW 7E LA14 8E NP 5W RH78 1 SR 9E

DA 2 LA15 8E NPS 13 RM 12 SR7 10

DD 15 LA16 8E NR 12 S 11 SR8 10

DE 6 LA17 8E NW 1 S18 6 SR9 9E

DG 8S LA18 8E OL 7E S19 11 SS 12

DH 10 LA19 8E OX 1 S32 6 ST 6

DH4 9E LA20 8E PA 14 S33 6 SW 1

DH5 9E LA21 8E PE 12 S34 11 SY 6

Continued on next page.


104

Postcode Zone Postcode Zone Postcode Zone Postcode Zone Postcode Zone

DH6 10 LA22 8E PE9 11 S40 6 SY14 7E

DL 10 LA23 8E PE10 11 S41 6 SY15 13

DN 11 LA24 7E PE11 11 S42 6 SY16 13

DT 3 LD 13 PE12 11 S43 6 SY17 13

DY 6 LE 6 PE13 12 S44 6 SY18 13

E 1 LL 7W PE20 11 S45 6 SY19 13

EC 1 LL23 13 PE21 11 S46 11 SY20 13

EH 15 LL24 13 PE22 11 S49 6 SY21 13

EH43 9S LL25 13 PE23 11 S50 11 SY22 13

EH44 9S LL26 13 PE24 11 SA 5W SY23 13

EH45 9S LL27 13 PE25 11 SA14 13 SY24 13

EH46 9S LL28 7W PE26 12 SA15 13 SY25 13

EH47 15 LL30 13 PH 15 SA16 13 SY26 6

EN 1 LL31 13 PH19 17 SA17 13 TA 5E

EN9 12 LL32 13 PH20 17 SA18 13 TD 9S

EX 4 LL33 13 PH21 17 SA19 13 TD12 9E

FK 14 LL34 13 PH22 17 SA20 13 TD13 9S

FY 7E LL35 13 PH23 17 SA21 5W TD15 9E

G 14 LL36 13 PH24 17 SA31 13 TD16 9S

GL 5E LL37 13 PH25 17 SA32 13 TF 6

GU 1 LL38 13 PH26 16 SA33 13 TN 2

GU11 3 LL39 13 PH27 15 SA34 13 TQ 4

GU12 3 LL40 13 PH30 17 SA35 13 TR 4

GU13 1 LL41 13 PH31 17 SA36 13 TS 10

GU14 3 LL42 13 PH32 17 SA37 13 TW 1

GU15 1 LL43 13 PH33 17 SA38 13 UB 1

GU28 2 LL44 13 PH34 17 SA39 13 W 1

GU29 2 LL45 13 PH35 17 SA40 13 WA 7E

GU30 3 LL46 13 PH36 17 SA41 13 WC 1

GU31 3 LL47 13 PH37 17 SA42 13 WD 1

GU32 3 LL48 13 PH38 17 SA43 13 WF 11

GU33 3 LL49 13 PH39 17 SA44 13 WN 7E

GU34 3 LL50 13 PH40 17 SA45 13 WR 6

GU35 3 LL51 13 PH41 17 SA46 13 WS 6

GU36 1 LL52 13 PH42 17 SA47 13 WV 6

GU46 3 LL53 13 PH43 17 SA48 13 YO 10

GU47 1 LL54 13 PH44 17 SA49 5W YO15 11

GU51 3 LL55 13 PH45 15 SA61 13 YO16 11

GU52 3 LL56 13 PH49 14 SA62 13 YO17 10

GU53 1 LL57 13 PH50 14 SA63 13 YO25 11

HA 1 LL58 13 PH51 15 SA64 13 YO26 10

HD 11 LL59 13 PL 4 SA65 13 ZE 20


105

Table 6.3-2: Average latitude and longitudes from properties within the designated MCS zones. These values were using to plot the path of the sun throughout the year in the assessment tool.

MCS Defined Zone Average Latitude Average Longitude

1 51.5 0

2 51 -0.5

3 51 2

4 50.5 4.1

5W 51.5 3.5

5E 51.2 2.8

6 52.5 1.8

7W 53 3

7E 53.5 2.6

8S 55 4

8E 54.5 3.2

9S 55.5 2.5

9E 55 1.9

10 54.5 1.5

11 53.5 1

12 52.5 -0.5

13 52.5 3.7

14 55.8 4.6

15 56.2 3.5

16 57.4 2.7

17 57.5 4.3

18 58 6.8

19 59 3.1

20 60.3 1.4

21 54.7 6.7


106

Figure 6.3-2: Example of MSC specified irradiance chart for zone 1 (Electrical Contractors Association, 2012).

0 5 10 15 20 25 30 35 40 45

0 828 828 828 828 828 828 828 828 828 828

1 835 835 835 835 835 835 834 834 833 833

2 843 843 843 842 842 841 841 840 839 838

3 850 850 850 849 849 848 847 846 845 843

4 857 857 857 856 855 854 853 852 850 848

5 864 864 864 863 862 861 859 857 855 853

6 871 871 870 869 868 867 865 863 861 858

7 878 877 877 876 874 873 871 868 866 862

8 884 884 883 882 880 879 876 873 870 867

9 890 890 889 888 886 884 882 878 875 871

10 896 896 895 894 892 890 887 883 880 875

11 902 902 901 900 898 895 892 888 884 879

12 908 908 907 905 903 900 897 893 888 883

13 914 913 912 910 908 905 901 897 892 887

14 919 919 917 916 913 910 906 901 896 890

15 924 924 922 920 918 914 910 905 900 894

16 929 929 927 925 922 919 914 909 903 897

17 934 933 932 930 927 923 918 913 907 900

18 938 938 936 934 931 927 922 917 910 903

19 943 942 941 938 935 931 926 920 913 906

20 947 946 945 942 939 935 929 923 916 908

21 951 950 949 946 943 938 933 926 919 911

22 954 954 952 950 946 941 936 929 922 913

23 958 957 956 953 949 944 939 932 924 915

24 961 961 959 956 952 947 941 934 926 917

25 964 964 962 959 955 950 944 937 928 919

26 967 967 965 962 958 953 946 939 930 921

27 970 969 968 965 960 955 948 941 932 922

28 972 972 970 967 962 957 950 942 933 923

29 975 974 972 969 964 959 952 944 935 924

30 977 976 974 971 966 960 953 945 936 925

31 979 978 976 973 968 962 955 946 937 926

32 980 979 977 974 969 963 956 947 937 926

33 982 981 979 975 970 964 957 948 938 927

34 983 982 980 976 971 965 957 948 938 927

35 984 983 981 977 972 966 958 949 938 927

36 984 984 981 978 973 966 958 949 938 927

37 985 984 982 978 973 966 958 949 938 926

38 985 984 982 978 973 966 958 949 938 925

39 985 984 982 978 973 966 958 948 937 925

40 985 984 982 978 973 966 957 947 936 924

41 984 984 981 977 972 965 956 946 935 922

42 984 983 981 977 971 964 955 945 934 921

43 983 982 980 976 970 963 954 944 932 919

44 982 981 979 975 969 962 953 943 931 918

45 980 980 977 973 967 960 951 941 929 916

Orientation (˚)

Incl

inat

ion

(˚)


107

Table 6.3-3: Diffuse/Global irradiation ratios used to calculate loss of incident energy at different shading levels (Muneer, 2004).

Station Latitude (˚) Diffuse/Global Irradiation Ratio

Jersey 49.2 0.51

East Hampstead 51.4 0.59

London 51.5 0.58

Aberporth 52.1 0.56

Cambridge 52.3 0.60

Aldergrove 54.6 0.61

Eskdalemuir 55.3 0.61

Lerwick 60.1 0.63

Table 6.3-4: Pricing structure prices p/kWh and quarter meter charges used to calculate pricing structure comparisons.

Pricing Structure

Automatic Meter Reading Units (p/kWh)

AMR (£/quarter)

Non-AMR (£/quarter)

All

Day

Nigh

t

We

ekd

ays

Even

ing

We

eken

d

Days

Unrestricted 26.84 26.84 9.54 - - - - -

Day/Night 26.84 26.84 - 9.99 6.02 - - -

Evening/Weekend 35.2 16.6 - - - 10.02 8.53 8.53

Day/Night/Evening/Weekend 35.1 16.5 - - 6.17 10.8 9.7 9.7


108

6.4 Alternative Calculators Graphs

Figure 6.4-1: Financial results graph taken from the Energy Saving Trust (2016).

Figure 6.4-2: Generation and financial graphs taken from Solar Guide (2015).


109

Figure 6.4-3: Financial results graph taken from Solar World UK (Solar World UK, 2016).

Figure 6.4-4: Financial results graph taken from Solar Panels UK (Solar Panels UK, 2016).


110

6.5 Case Study Results Print-outs

6.5.1 Commercial Property

3 Chapel Court

CH2 4BT

Tel : 01244 399399

[email protected]

System Specifications

Postcode: CH2 4DW EPC Rating: C

Inclination: 45 ˚ Orientation: 25 ˚

Area Classification: Countryside Shading Factor: Light Shading (10%-20%)

System Peak Power: 4 kWp Property Classification: Commercial

Annual Electricity Usage: 3800 kWh Electricity Distributor: UK Average

Year of Installation: 2016 Quote Price: £6,171

Assumptions Key Results

System Losses: 15% Average Annual Generation: 2475 kWh

Maintenance Costs: 1.0% Actual Export Ratio: 5.9%

Inflation: 3.0% Lifetime Profits: £3,826

Degradation Rate: 0.5% Return on Investment: 3.1%

System Lifetime (years): 20 Payback Period: 10.1 Years

MCS Irradiance (kWh/kWp): 845 CO2 Lifetime Savings: 24255 kg

Energy Payback Period: 5.6 Years

19.9

Bespoke Solar Assessment

CH2 4DW

58387

596

0

100

200

300

400

kWh

/mo

nth

Yearly Energy Curves

Usage Generation

0

0.1

0.2

0.3

0.4

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

kWh

/hal

f ho

ur

Average Daily Energy Curves

Usage Generation

£60

£109

£443

Annual Revenue Breakdown

Annual Export PaymentAnnual FiT PaymentAnnual Self-Consumption Sav ings

0

2000

4000

6000

8000

10000

12000

0 5 10 15 20

Cas

h F

low

(£)

Years after installation

Net Profit

System Cost

× 1000

Thousand Cups of Tea over System Lifetime

Acres of Forest Annual CO2

Absorbtion Equivalent

Miles Offsetover System Lifetime


111

6.5.2 Educational Property

3 Chapel Court

CH2 4BT

Tel : 01244 399399

[email protected]





System Peak Power: 4 kWp Property Classification: Educational











19.9


CH2 4DW

58387

596

0

100

200

300

400

500

kWh

/mo

nth


Usage Generation

0

0.1

0.2

0.3

0.4

0.5

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

kWh

/hal

f ho

ur


Usage Generation

£60

£109

£417



0

2000

4000

6000

8000

10000

12000

0 5 10 15 20

Cas

h F

low

(£)


Net Profit

System Cost

× 1000






112

6.5.3 Industrial Property

3 Chapel Court

CH2 4BT

Tel : 01244 399399

[email protected]





System Peak Power: 4 kWp Property Classification: Industrial











19.9


CH2 4DW

58387

596

0

100

200

300

400

500

kWh

/mo

nth


Usage Generation

0

0.1

0.2

0.3

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

kWh

/hal

f ho

ur


Usage Generation

£60

£109

£441



0

2000

4000

6000

8000

10000

12000

0 5 10 15 20

Cas

h F

low

(£)


Net Profit

System Cost

× 1000






113

6.5.4 Light Commercial

3 Chapel Court

CH2 4BT

Tel : 01244 399399

[email protected]





System Peak Power: 4 kWp Property Classification: Light Commercial











19.9


CH2 4DW

58387

596

0

100

200

300

400

500

kWh

/mo

nth


Usage Generation

0

0.1

0.2

0.3

0.4

0.5

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

kWh

/hal

f ho

ur


Usage Generation

£60

£109

£415



0

2000

4000

6000

8000

10000

12000

0 5 10 15 20

Cas

h F

low

(£)


Net Profit

System Cost

× 1000






114

6.5.5 Residential

3 Chapel Court

CH2 4BT

Tel : 01244 399399

[email protected]





System Peak Power: 4 kWp Property Classification: Residential











19.9


CH2 4DW

58387

596

0

100

200

300

400

500

kWh

/mo

nth


Usage Generation

0

0.1

0.2

0.3

0.4

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

kWh

/hal

f ho

ur


Usage Generation

£60

£109£382



0

2000

4000

6000

8000

10000

12000

0 5 10 15 20

Cas

h F

low

(£)


Net Profit

System Cost

× 1000






115

6.5.6 Social

3 Chapel Court

CH2 4BT

Tel : 01244 399399

[email protected]





System Peak Power: 4 kWp Property Classification: Social











19.9


CH2 4DW

58387

596

0

100

200

300

400

500

kWh

/mo

nth


Usage Generation

0

0.1

0.2

0.3

0.4

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

kWh

/hal

f ho

ur


Usage Generation

£60

£109

£414



0

2000

4000

6000

8000

10000

12000

0 5 10 15 20

Cas

h F

low

(£)


Net Profit

System Cost

× 1000





development of a solar pv energy assessment tool for eg-audit ltd

Documents