NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.

Understanding Moisture Ingress Rates in PV Modules

Barrier Technologies Workshop

Michael Kempe

September 19, 2012

2

Introduction

• Many electronic devices are sensitive to moisture (e.g. OLEDs, OPV, CIGS, CdTe).

• For some applications we would like to have a transparent flexible moisture barrier material. o Ease of application.

o Weight limits for some structures.

o Here I will explain the theory behind why such low permeation is necessary and how to prevent moisture ingress.

3

Outline

• Permeation through barrier films. 25 y requirements.

• Diffusion from the module edges.

• Edge seals

4

How Does WVTR Relate to Total Permeation?

This depicts how much water would get in if it was consumed immediately.

0.01

0.1

1

10

100

1000

10000

100000

1000000

1.00E-061.00E-051.00E-041.00E-031.00E-021.00E-011.00E+001.00E+011.00E+02

Wat

er

Laye

r Th

ickn

ess

(µ

m)

WVTR (g/m2/day)

Water Thickness Permeated in 25 y

10-6 10-5 10-4 10-3 10-2 10-1 100 101 102

Thin Film LayerThicknesses

WVTR

PET(0.10 mm)

PCTFE(0.023 mm)

PDMS(1 mm)

PEN(0.10 mm)

5

Moisture in Breathable Front-Sheets

Moisture Permation Assumptions:

(1) CEVA not a function of position X (i.e. DEVA>>DB)

(2) WVTR is proportional to DCB=C(0)-C(lB)

BBB

EVASatE

SatBEVA lCClC

WVTR

dt

dC 0

.,

,

H2O

PV Device

Barrier Film

Encapsulant

Dielectric films Metal Foil film

6

Time Constant for Water Ingress

ingressmoistureofRate

holdcanEVAwaterofAmout

WVTR

lC

SatB

EVAEVASat693.0693.0

,

,

21

H2O

PV Device

Barrier Film

Encapsulant

Dielectric films Metal Foil film

EVAEVASat

Sat

lC

tWVTR

o eCtC ,1 BBB

EVASatE

SatBEVA lCClC

WVTR

dt

dC 0

.,

,

7

Time Constant for Water Ingress

lEVA =0.46 mm, T=27 oC, CSat,EVA=0.0022 g/cm3

lPET=0.10 mm, lPEN=0.10 mm, lPCTFE=0.022 mm

PET PEN PCTFE

1/2 = 0.22 0.91 5.5 (day)

For 1/2=25 years need 0.8·10-4 g/m2/day

At 1·10-3 g/m2/day, the encapsulant will have a half

time of 1.9 y. This short time frame is insignificant

compared to a 25 y warranty.

SatB

EVAEVASat

WVTR

lC

,

,

21 693.0

8

Encapsulant Materials Structures

Si O

CH3

CH3

n

n

O

O

mn m

O O (Na, Zn)

RN

O

O

H n

OO HO OO

n m o

Ionomer

Thermoplastic Polyurethane (TPU)

Polyvinyl Butyral (PVB)

Ethylene Vinyl Acetate (EVA)

Polydimethylsiloxane (PDMS)

R

n m

R= -CH3 ,-(CH2)nCH3, others or multiple "R" groups.

Thermoplastic Polyolefin (TPO)

M. Kempe, “Overview of Scientific Issues Involved in Selection of Polymers for PV Applications”, 37th IEEE PVSC (2011).

SatB

EESat

WVTR

lC

,

,

21 693.0

9

Significantly More Adsorbent Polymers Exist

A 10X more water adsorbent polymer may reduce the barrier requirements by a factor of 10.

0 0.2 0.4 0.6 0.8 1

0

0.2

0.4

0.6

0.8

1

1.2

0.0%

0.0%

0.1%

1.0%

10.0%

0.0027 0.0029 0.0031 0.0033 0.0035 0.0037

Temperature (⁰C)

Solu

bil

ity (

g/c

m3)

1/Temperature (1/K)

PVB #1

PVB #2

TPU #2

TPU #1

Ionomer #1

EVA

Ionomer #2

PDMS

TPO #2

Poly-α-Olefin #1

Poly-α-Olefin #2

10%

1%

0.1%

0.01%

0.001%

97.2 71.7 49.4 29.9 12.6 -2.89

10

EVA Allows Significant Moisture Ingress From Edges

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0 5 10 15 20

Distance X from edge (cm)

Dis

solv

ed

Wate

r (g

/cm

3) 1.109360481

2.109129888

3.110725775

4.114604761

5.114374168

6.114143575

7.113912982

8.113682389

9.113451796

10.1132212

12.11276002

12.4166406

EVA Glass

Glass

H2O

X=0 X=20

H2O

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Increasing

Months

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Finite element analysis using meteorological data from Miami Florida 2001

11

Alternative Encapsulants Slow Down Moisture Ingress

0 0.2 0.4 0.6 0.8 1

0

0.2

0.4

0.6

0.8

1

1.2

1.0E-10

1.0E-09

1.0E-08

1.0E-07

1.0E-06

1.0E-05

1.0E-04

0.0027 0.0029 0.0031 0.0033 0.0035 0.0037

Temperature (⁰C)

Dif

fusi

vit

y (

cm2/s

)

1/Temperature (1/K)

PDMS #1

Poly-α-Olefin #1

Poly-α-Olefin #2

EVA

TPO #2

TPU #1

TPU #2

PVB

PVB #2

Ionomer #2

Ionomer #1

97.2 71.7 49.4 29.9 12.6 -2.89

10-4

10-5

10-6

10-7

10-8

10-9

10-10

Lower diffusivity can reduce ingress rates from the side by two orders of magnitude.

12

Determining Moisture Ingress Distance From Edges

M.D. Kempe, A. A. Dameron, M.O. Reese, to be submitted to Progress in Photovoltaics (2012)

PDMS EVA TPO Ionomer #1 PDMS EVA TPO Ionomer #1Munich, Germany 19.5 23.3 28.3 29.2 30.3 38.2 47.7 49.2Denver, Colorado 25.8 30.1 35.3 36.2 40.2 48.4 56.7 57.9Miami, Florida 33.8 35.6 37.9 38.3 44.4 48.9 54.3 55.2Albuquerque, New Mexico 30.6 34.6 39.4 40.1 46.4 54.0 61.3 62.4Bangkok, Thialand 38.0 39.7 42.0 42.4 48.9 53.4 58.9 59.8Phoenix, Arizona 40.6 40.6 48.7 44.3 56.6 63.8 71.1 72.1Riyadh, Saudi Arabia 42.3 45.8 50.1 50.8 58.0 65.0 72.2 73.3

Rack Mounted Insulated Back

Diffusivity Weighted Average Module Temperature

Effective Temperature (⁰C)

1

10

100

1,000

0.0027 0.0029 0.0031 0.0033 0.0035

Wid

th o

f ed

ge

se

al

(cm

)

1/K

25 year 5% Breakthrough Width

PDMS

Poly-α-olefin

EVA

TPO

TPU #1

TPU #2

PVB #1

PVB #2

Ionomer #2

Ionomer #185⁰C 60⁰C 20⁰C

Distance from edge which after 25 years will be at 5% of the equilibrium concentration.

EaPDMS=27 kJ/mol EaEVA=38 kJ/mol EaTPO=53 kJ/mol

EaIonomer#1=56 kJ/mol

13

Test Sample Designed to Mimic Module Edge

Seal Encapsulant

Glass

Glass

H2O

w

50 mmGlass (3.18 mm)Polymer Film (~0.5 mm)Ca (100 nm)Glass (3.18 mm)

H2O

Ca + 2 H20 → Ca(OH)2 + H2

Module Edge Test Sample

n

Polyisobutylene

Desiccant (e.g. CaO or molecular sieves,

Preferably type 3A Molecular Steve that does not absorb O2 and N2)

Edge Seals +

14

Oxidation of Ca Indicates Moisture Ingress

Ca + 2 H20 → Ca(OH)2 + H2

Mirror-Like → Transparent

H2O

H2O

50 mm

Unexposed 1500 h, 85⁰C, 85% RH

15

Moisture Ingress Rate Governed by Diffusion

15

H2O

H2O

Moisture ingress measured at 45ºC and 85ºC, with RH held at 85%, and at lower levels using saturated salt solutions of LiCl, MgCl, or NaNO3.

45 85

(⁰C) (⁰C)

NaNO₃ 67% 59%

MgCl 31% 25%

LiCl 11% 10%

RH (%)

16

Edge Seal Modeling

%100

%RHeSS kT

Ea

om

s

kT

Ea

oeff

D

eDD

Mobile phase water absorption is split between the

polymer matrix and the mineral components.

Assume linearity with relative humidity.

Mobile phase water diffusivity is an effective diffusivity.

This accounts for a rapid equilibration between adsorbed

and dissolved water.

A non-reversible

reaction with water

that immobilizes the

water. OHR

2

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0 20 40 60 80 100

K (

cm/h

1/2

)

% RH

85⁰C

45⁰C

PIB #2

Values for the 5 constants were found from absorption measurements and a fit to the data.

17

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 2 4 6 8 10

Ingr

ess

De

pth

(cm

)

Time (Years)

Denver Colorado, PIB #2

Insulated Back, Glass/Polymer

Close Roof, Glass/Glass

Open Rack, glass/glass

Open Rack, Glass/Polymer

Square Root Relation Works to Longer Times

Denver Colorado

Used TMY3 Data and Temperature estimate methods from to King et al.

tKX

18

Results for Different Climates

Glass/Polymer Modules.

A sensitivity analysis gave about ±15% on K and Width, and ±30% on 25 yr equivalent time.

0.33

47

0.16

5

0.047 (cm/h1/2) (cm) (h)

0.00087 0.45 1,100

0.00103 0.47 1,300

0.00096 0.50 1,400

0.00107 0.51 1,500

0.00102 0.50 1,400

0.00124 0.53 1,600

0.00128 0.60 2,000

0.00153 0.63 2,300

0.00199 0.90 4,700

0.00225 0.95 5,200

0.00228 1.02 6,000

0.00258 1.07 6,700

Miami, FlordiaOpen Rack

Insulated Back

Open RackBangkok, Thailand

Insulated Back

Riyadh, Saudi ArabiaOpen Rack

Insulated Back

Open RackPhoenix, Arizona

Insulated Back

Denver, ColoradoOpen Rack

Open Rack

Insulated Back

Munich, GermanyInsulated Back

Do (cm2/s)=

Modeled K Modeled 25 y

required width

Modeled 25 y

equivalent time

at 85⁰C/85% RH

EaD (kJ/mol)=

So (g/cm3)=

EaS (kJ/mol)=

Reactive Ca absorption (g/cm3)=

19

What edge seal parameters are important?

M. D. Kempe, A. A. Dameron, T. J. Moricone, M.O. Reese, “Evaluation and Modeling of Edge-Seal Materials for Photovoltaic Appilcations, 35th IEEE PVSC, Honolulu, HI (2010).

1. Adhesion is the most important parameter. a) Must be maintained after environmental exposure.

b) Residual stress in glass will affect adhesion.

c) Material may expand as it absorbs water.

d) Good surface preparation is necessary.

2. Breakthrough time is the next most important. a) The 12 mm edge delete perimeter should be wide enough to keep

moisture out.

3. Module mounting configuration is not important. a) Hotter installations tend to dry out the module partially countering the

effects of increased diffusivity.

4. The steady state transmission is less important. a) The amount of permeate is very low.

b) Ideally one will not reach steady state.

20

Conclusions

• An ingress half time of 25 years is needed. For typical barriers and encapsulants, a WVTR of less than 0.8·10-4 g/m2/day or better is needed.

• High solubility encapsulants may decrease the barrier needs to as low as 1·10-3 g/m2/day or better.

• With impermeable front and backsheets, very low diffusivity polymers can limit moisture ingress to a few cm from the edges.

• A PIB based edge seal width of 1 cm should be able to prevent moisture ingress.

21

Acknowledgements

Sarah Kurtz John Wohlgemuth

David Miller Joshua Martin

Arrelaine Dameron Matthew Reese Tom Morricone

Dhananjay Panchagade

This work was supported by the U.S. Department of Energy under Contract No. DE-AC36-08-GO28308 with the National Renewable Energy Laboratory.