modelling of isolated solar pv households with battery...
TRANSCRIPT
Monash
University
Global
Innovation
Modelling of Isolated Solar PV
Households with Battery Energy
Storage
Dr Ross Gawler
Senior Research Fellow
Monash University
Monash
University
Global
Innovation
Introduction
Electrification of remote villages remains an important strategy
in developing countries
Solar panels and batteries are reducing in cost and competing
with grid supplied energy in niche markets
– May obviate network extension to remote areas generally
Research Questions for solar/battery technology:
– Optimal deployment to supply remote areas?
– Clustering of buildings through microgrids?
• Maximum distance for connecting small households could
be interconnected?
The focus is on Indonesia for the Australia Indonesia Centre
Monash
University
Global
Innovation
The Design Formulation - Demand
Assume no metered demand for electricity
Four components versus household income:
– Household activity profile – randomised for each day
– Appliance ownership
– Appliance power characteristics (standby and maximum)
– Energy consumption for various activities
Link appliances to activities
Stochastic demand model for each activity and appliances
Some loads are deferrable for up to 8 hours if supply becomes
available during this period
Unserved energy and deferred energy valued:
– US$5.00/kWh for unserved
– US$2.50/kWh for deferred energy
Monash
University
Global
Innovation
The Design Formulation - Supply
Choose battery and solar panel options:
– 320 W panels
– Lead-carbon batteries
– Separate infrastructure cost for space and electrical equipment
Stochastic solar model based on random daily energy and half-
hour sampling to allocate daily energy
Simulation of alternative combinations of panels and batteries
– optimise panels and batteries for each income level for a single
household of each size
– Include cost of unserved and deferred energy to optimise
reliability
Stochastic simulation of two connected households to assess
value of interconnection
– Savings in panels and batteries available due to interconnection
Monash
University
Global
Innovation
Electricity consumption versus income
Income and appliance ownership – key drivers of demand
AS Permana, Sept 2008 showed a relationship between
income and energy usage (including LPG)
Average Indonesia usage about 130 kWh/month (PLN)
Average energy demand also related to settlement size
(CastleRock)
100
1000
10000
100 1000
Ho
use
ho
ld e
ne
rgy
kW
h p
er
mo
nth
Average monthly Income USD
Household Consumption versus Income
Data
Fit
100
110
120
130
140
150
160
170
180
2016 2017 2018 2019 2020 2021 2022
kW
h/c
usto
mer/
month
Forecast Year
Forecast monthly sales per residential customer
Java-Bali Sumatra East Indonesia Indonesia
Monash
University
Global
Innovation
Energy use for applications
Sorapipatana 2016 showed energy use by household activities
at four different consumption levels
– Interpolate and extrapolate to range 10 – 200 kWh/month
0
10
20
30
40
50
60
70
80
90
100.00 120.00 140.00 160.00 180.00 200.00 220.00 240.00 260.00
Co
mp
on
en
t C
on
sum
pti
on
kW
h/m
on
th
Total Consumption kWh/month
Usage Components
Cooking Entertainment Laundry Water supply Air-Cond
Lighting Other Refrigeration Fan
0%
5%
10%
15%
20%
25%
30%
35%
40%
0 50 100 150 200 250
Co
mp
on
en
t C
on
sum
pti
on
kW
h/m
on
th
Total Consumption kWh/month
% Smoothed Usage Components
Cooking Entertainment Laundry & Housework
Water supply Lighting Other
Air-conditioning Refrigeration Fan
Monash
University
Global
Innovation
Activity Modelling
Created seven activities for individuals and households
– Sleeping– Cooking– Eating– Personal Care– Laundry and housework– Entertainment– Absence
Allow one core activity for the household
– individuals not considered Model household activity as a
random activity with different probabilities over the day for
– Work days– Non-work days
Based on modelling by (Wilke 2013)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Activity Profile - Work Day
Sleeping Cooking Laundry & Housework
Entertainment Personal Care Absence
Eating
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Activity Profile - Non-work Day
Sleeping Cooking Laundry & Housework
Entertainment Personal Care Absence
Eating
Monash
University
Global
Innovation
Random sampling of energy use
Derive activities and energy use profiles as random processes
126 kWh/month
0.000
0.050
0.100
0.150
0.200
0.250
0.300
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Ave
rage
Dem
and
kW
Hour of the Work Day
Workday Expected Energy Power Use by Activity
Sleeping Cooking Eating
Laundry & Housework Entertainment Personal Care
Absence
0.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Ave
rage
Dem
and
kW
Hour of the Work Day
Non-workday Expected Energy Power Use by Activity
Sleeping Cooking Eating
Laundry & Housework Entertainment Personal Care
Absence
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Dem
and
kW
Hour of the Day
Expected Power Work day
Maximum Power byTime of Day
Expected Power by Timeof Day
0.0
0.5
1.0
1.5
2.0
2.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Dem
and
kW
Hour of the Day
Expected Power Non-work day
Maximum Power byTime of Day
Expected Power by Timeof Day
Monash
University
Global
Innovation
Solar Energy
Solar energy is based upon a solar insolation model which
takes account of
– Latitude and longitude of the location
– Orientation of the panels
– Rating of the panels
– Allowance for shading at start and end of day
– Allowance for direct and diffuse insolation
Initial model based on public data from an 11 kW Jakarta
Installation from pvoutput.org
– Fitted the model parameters to match the part-year daily energy
data
Monash
University
Global
Innovation
Modelling of solar energy
Model daily energy as an auto-regressive model (data fitted)
Model half-hour energy as an auto-correlated profile (50%) to
match the sampled daily energy within maximum and minimum
daily profile by time of the year
0
10
20
30
40
50
60
Daily Target Energy and Half-hour Simulation
Target Daily Energy Simulated Daily Energy
0.00
2.00
4.00
6.00
8.00
10.00
12.00
1213141516171819202122232425262728293031323334353637383940
Sampled Hourly profiles
12-Sep-17 13-Sep-17 14-Sep-17 15-Sep-17
16-Sep-17 17-Sep-17 18-Sep-17
Monash
University
Global
Innovation
Batteries
Initial model based on Narada lead-carbon batteries
Three sizes were selected based on data provided by PT
Solar Power Indonesia
Units Small Medium Large
Gross Capacity kWh 1.44 3.6 7.2
Usable Capacity kWh 0.96 2.4 4.8
Maximum Power kW 0.36 0.9 1.8
Cycle efficiency % 90.99% 91.98% 92.95%
Unit Cost $/kW $775 $620 $620
Infrastructure Cost $ $922 $1,126 $1,283
Technical Life Years 10 10 10
Annual capacity
degradation % 3.3% 3.3% 3.3%
Monash
University
Global
Innovation
Single House Dispatch MethodRandom Solar Power Random Load
Compare half-hourly
Surplus Deficit
Shed or Defer Load
Charge Batteries Discharge Batteries
Unused Solar Energy Unserved Load
Unserved Energy CostInfrastructure Cost +Total Cost =
State of
Charge of
Batteries
Recover Deferred Load
Choose S
ola
r P
anels
and B
att
eries
Monash
University
Global
Innovation
Observations from Single House Analysis
Economic issues
– Costs exceed income below 20 kWh/month
– Costs are about 63% of income at average national
consumption of 130 kWh/month
• Marked costs reductions needed to support complete
remote electrification unless electricity can promote
increase of income (cost and utility to be considered)
Technical guide:
– Battery power and solar power capacity are closely aligned
– Battery storage capacity is 1.5 to 2 times the average daily
energy demand
– Battery unit size increases with demand level over the range
Monash
University
Global
Innovation
Single House Solutions
0
2
4
6
8
10
12
10.0 13.8 18.9 41.4 91.0 200.0
Num
ber
of
Units
Average energy demand kWh/month
Single House - Units for Optimal Design
Panels Small Medium Large
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
10.0 13.8 18.9 41.4 91.0 200.0
kW
Average energy demand kWh/month
Peak Supply and Demand
Battery Power kW Solar Power kW Peak Demand kW
0
2
4
6
8
10
12
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00
Tota
l U
sable
Battery
Capacity
kW
h
Average Daily Demand kWh/day
Battery Capacity versus Daily Demand
$0.00
$0.20
$0.40
$0.60
$0.80
$1.00
$1.20
$1.40
$1.60
$0
$20
$40
$60
$80
$100
$120
$140
$160
$180
10.0 13.8 18.9 41.4 91.0 200.0
US
D/k
Wh
US
D/m
onth
Average energy demand kWh/month
Cost, Income and Price
Average Cost $/month Income $/month Average Cost $/kWh
Monash
University
Global
Innovation
Two- House Interconnection Value
Single House Simulation
Optimal Design vs Demand
Solar Model Appliances and ActivitiesBattery Model
Two House Simulation
Optimal Design
Compare Costs
Maximum DistanceConnection Cost
Monash
University
Global
Innovation
Maximum Distance – to 120 meters
10.013.8
18.941.4
91.0
200.0
0
20
40
60
80
100
120
140
10.0
13.8
18.9
41.4
91.0
200.0
kWh Monthly Consumption
meters
kWh Monthly Consumption
Maximum Connection Distance
0-20 20-40 40-60 60-80 80-100 100-120 120-140
200.0
91.0
41.4
18.9
13.8
10.010.0 13.8 18.9 41.4 91.0 200.0
KWH MONTHLY CONSUMPTION
METERS
KW
H M
ON
THLY
C
ON
SUM
PTI
ON
Maximum Connection Distance
0-20 20-40 40-60 60-80 80-100 100-120 120-140
Monash
University
Global
Innovation
Different solar panel orientation – 190m meters
10.013.8
18.941.4
91.0
200.0
0
50
100
150
200
10.0
13.8
18.9
41.4
91.0
200.0
kWh Monthly Consumption (Original)
meters
kWh Monthly Consumption (Alternative)
Maximum Connection Distance
0-50 50-100 100-150 150-200
200.0
91.0
41.4
18.9
13.8
10.010.0 13.8 18.9 41.4 91.0 200.0
KWH MONTHLY CONSUMPTION
METERS
KW
H M
ON
THLY
C
ON
SUM
PTI
ON
Maximum Connection Distance
0-25 25-50 50-75 75-100 100-125 125-150 150-175 175-200
Monash
University
Global
Innovation
Maximum Distance
Based on these results: houses are worth interconnecting up to
120 m distant...
– If the patterns of demand are disparate in volume and timing
– If the consumption is below the efficient scale of the available
solar panels and batteries
– If no constraints on solar panel installation and same roof
orientation
If roof orientation is dissimilar, then maximum interconnection
distance may be up to 190 m
– Design optimisation is more complex to match solar panel
location to patterns of prospective demand
Monash
University
Global
Innovation
Other issues
Further considerations not studied…
– Limited roof space constraining solar capacity
– Limited space for security of battery facility
– Lower cost connection at DC voltage for shorter distances?
– Connection of residential and community buildings
– Economies of scale with centralised solar installation supplying
many houses
Monash
University
Global
Innovation
Next Steps
Seek a software platform to enable more complicated networks
with multiple buildings to be solved
– PLEXOS
– Minizinc
– Bespoke programming
Access survey data on household activities, buildings and
energy consumption for a prospective village project and test
whether these conclusions are robust.
Formulate a microgrid planning method that can quickly assess
– Interconnectness among premises
– Aggregate costs for regional economic planning
– Value for micro-economic development
– Value for future interconnection with main grid
Monash
University
Global
Innovation
Contact details
Dr Ross Gawler
Senior Research Fellow, Monash University
+61 3 9504 8373
+61 419 890 723
Monash
University
Global
Innovation
Modelling of Isolated Solar PV
Households with Battery Energy
Storage