Intersolar Middle East September 25-26, 2017 Novel soiling - - PowerPoint PPT Presentation

Novel soiling measurement technology

Xander van Mechelen, CTO

Intersolar Middle East

September 25-26, 2017

It is our mission to be a global, leading manufacturer of instruments that measure climate properties for the long term benefit of people.

Insight in the basics of Soiling

1.1 PV performance dependencies 1.2 Dust characteristics 1.3 Effects on PV panels 1.4 Existing solutions

New Instrument development

2.1 Novel technology 2.2 Testing and characterisation 2.3 Application 2.4 Conclusions

Three main parameters

(besides hardware issues)

Irradiance Temperature Soiling

1 3 2

High quality, good understanding Qualitative opinions & basic or no measurements

1 + = 2 = 3

PV performance variations

Soiling the big unknown

Three main parameters

(besides hardware issues)

Soiling is the only parameter that you can influence!

Irradiance Temperature Soiling

1 3 2

High quality, good understanding Qualitative opinions & basic or no measurements

1 + = 2 = 3

PV performance variations

Soiling the big unknown

Soiling losses are significant most attractive PV locations are troubled by soiling

Soiling loss location dependent, 0 to 2% per day Within one week, power loss can be >10% Rain events reduce soiling losses Current best practice: accepted loss 4.4% Business case for cleaning cost vs loss is location dependent

Soiling accumulation over days Soiling accumulation loss per day

Average of annual solar insolation [kWh/m2/day] 5.92 5.61 5.40 5.08 5.77 5.42 5.16 4.65 0.5 1 1.5

Libya Abu Dhabi, UAE Kuwait, Kuwait Mountain View, CA, USA Riyadh, Saudi Arabia Limassol, Cyprus Ogbomoso, Nigeria Dhaka, Bangladesh

2 5 1 51 101 201 151 251 301 351 10 15 20 25 30 Day Daily output power loss increase [%]

Sayyah et al. Solar energy (2014) Zorrilla-Casanova et al. (2011)

Loss (%) Rainfall (mm) 20 40 60 80 100 120 140 Daily losses Rainfall

About soiling and dust Sources of soiling

dust • plant products • soot • salt • bird droppings

1000-500 500-250 250-125 125-63 63-31 31-16 16-8 8-4 < 4 0.00 0.00 0.82 4.78 8.16 16.47 23.82 20.19 25.75 Coarse grained Medium grained Fine grained Very fine grained Coarse silt Medium silt Fine silt Very fine silt Clay D(µm) % of the total sample Grain type Dust grain distribution and sediment types Quartz Calcite Albite, calcian, ordered Dolomite Muscovite Palygorskite Lizardite-1T Kaolinite 1Md SiO2 CaCO3 (Na, Ca) Al (Si, Al)3 O8 CaMg (CO3)2 KAl3Si3O10(OH)2 Mg5 (Si, Al)8 O20 (OH)2 8H2O Mg3Si2 O5 (OH)4 Al2Si2O5 (OH)4 Compound name Chemical formula Dust sample material composition obtained using XRD analysis

Zone 1 5.2 - 8.1 8.1 - 12 Zone 2 12 - 19 19 - 30 Zone 3 30 - 44 44 - 65 Zone 4 65 - 96 96 - 142

Dust intensity around world, based on M. Maghami et al. 2016

Humidity and dew can be a determining factor in soiling impact; ‘stickiness’ of dust Local grain size determins optimimum PV tilt angle

• trade off power production vs soiling accumulation

Acculumation speed of dust can be a sort of a local ‘constant’ and maintenance can be more or less planned Timing of cleaning practices is local knowledge (season, time of day) Evolving technologies for cleaning robots as an intergral part of PV plant design

Qualitative insights for dust practices

Local insights and inventions of best practices

Existing solutions

Inspiration for a new product

All based on 2 panel comparison: € 9,000 to € 30,000

Inspiration: can we... prevent daily cleaning/maintenance? avoid moving parts/fluids? provide multiple sensors across the plant for a similar budget? measure in more representative wind conditions? improve ease of installation?

Moroni & Partners UKC DDSolar Atonometrics Soiling Campbell

DustIQ principle

Optical Soiling Measurement (OSM) Technology

Measured from the inside

Reflection → Transmission loss → Soiling loss

LED Photodiode ‘clean’ signal ‘soiled’ signal LED Photodiode

DustIQ signal response

response curve

Signal strength at different soiling rates

Netto loss [mV] Transmission loss [%]

Photodiode LED Pyranometer

Sun or lamp

Soiling X% Soiling Y% Signal X1 Signal Y1 Signal X2 Signal Y2 Glass plate Situation X Situation Y

DustIQ Curves

Different curves for different colors of dust

Test dust vs Abu Dhabi and Jordan

black dust brown dust white dust Abu Dhabi Jordan

Transmission loss [%] Netto loss [mV]

DustIQ Characterization

A B C D E

Designed for Inhomogeneity

diameter more than enough

3 µm 10 µm 20 µm 40 µm

Transmission loss DustIQ signal [mV]

5

• 5
• 10
• 15
• 20
• 25
• 30

10 15 20 25

No grain size effects

Field application

measurement at sepearate locations position in the middle of the solar panels position at the top of a solar panel

Field application Understanding of output

Transmission Loss(dust type) [%] = Reflected signal x Manufacturer calibration x Field calibration(dust type) Soiling Ratio(dust type) [%] = 100% - Transmission Loss

Determine your dust color Field calibration Install next to panel Mounting

Three main parameters PR by two sensors

Irradiance Temperature Soiling Pyranometer DustIQ DustIQ

1 3 2

Soiling Ratio 95 - 90% Soiling Ratio 90 - 80% Soiling Ratio 80 - 50% Measurement interval Ambient temperature range Weight Instrument dimensions +/- 1 to 2% +/- 2 to 4% +/- 5 to 10% 1 min, IEC61724 compliant

• 20 to +60°C

5 kg 990 x 160 x 40 mm Soiling Ratio 100 - 95% +/- 0.5 to 1% Technical specifications DustIQ

Results

DustIQ specification

Patents applied for in Europe & China International coverage Self-calibrating for dust color/LED/photodiode changes, etc.

Near future optimization

Compensate for soiling ratio variations during the day

Most accurate around noon and during the night (no dew) Dependence on solar inclination angle Effect holds for both DustIQ and 2 panel comparison methods

5 10 15 20 25 06:00 08:24 10:48 13:12 15:36 18:00 20:24 22:48

TL (%) Local time

Conclusions

solar noon solar noon +2 hr solar noon - 2hr

Near future optimization

Compensate for soiling ratio variations during the day

Soiling Ratio [SR]

0.83 0.84 0.85 0.86 0.87 10:33 11:45 12:57 14:09

Variation SR with the angle of incidence due to artificial soiling at Delft for 14% Tloss

Time Time

15:21 16:33

Photo of roof PV setup with artificially soiled PV panels and SR as function of varying solar angle over the course of a day

Near future optimization

Compensate for soiling ratio variations during the day

TL (%) solar noon soiling ratio

20 40

22% 18% 14% 9% 4% 22% 18% 14% 9% 4%

60 80 100 40 3 5 30 25 20 15 10 5

Irradiance losses for solar angle with equation fit

SA (degrees)

∆ 3% ∆ 7%

SR(out) can be different from instantaneous SR and can be compensated by calculations to make DustIQ accurate 24/7

Data to end-user Afer processing

• at monitoring company with NDA
• via Kipp & Zonen server

Data to SCADA

1 to 3% uncertainty 4 to 6% uncertainty

Near Future

Timing optimization of your cleaning robots

DustIQ to detect dew Clean when dew is just gone, and before the sun ‘bakes’ the dust firmly to the PV modules DustIQ to trigger the cleaning robots in this timeslot

Introduced novel soiling measurement technology We can Performance demonstrated high sensitivity/precision signals linear grain size independent color characterisation dew detection Field application and shared data will improve future usage test locations in Delf and Spain. Planned: Arizona and Qatar First delivery mid December 2017

Conclusions

prevent daily cleaning/maintenance? avoid moving parts/fluids? provide multiple sensors across the plant for a similar budget? measure in more representative wind conditions? improve ease of installation?

For more information please mail donald.van.velsen@kippzonen.com

Thank you for your attention