Technical Assumptions Used in PV Financial Models: Review of Current Practices and Recommendations

IEA INTERNATIONAL ENERGY AGENCY
PHOTOVOLTAIC POWER SYSTEMS PROGRAMME
Technical Assumptions Used in PV
Financial Models: Review of Current
Practices and Recommendations
Mauricio Richter (3E), Jan Vedde (SiCon),
Mike Green (M.G.Lightning Electrical Engineering),
Ulrike Jahn, Magnus Herz (TÜV Rheinland)
IEA PVPS TASK 13 WEBINAR, 5 JULY 2017

Outline
• Mitigating the Risks Inherent with Technical Assumptions Used in PV
Financial Models (Mauricio Richter)
• Mitigating Financial Risks in a PV Investment (Mike Green)

Mitigating the Risks Inherent with
Financial Models
Mauricio Richter (3E), Jan Vedde (SiCon),
Ulrike Jahn, Magnus Herz (TÜV Rheinland)
IEA PVPS TASK 13 WEBINAR, 5 JULY 2017

Outline
• Introduction
• Technical Assumptions used in PV Financial Models - Weaknesses
& Recommendations:
– PV Plant Design
– Procurement Process
– Plant Construction
– Plant Acceptance
– O&M
• Long-Term Yield Estimates & Their Level of Confidence
• Risk Quantification and Economic Impact
• Conclusions

Introduction
• Survey of current practices (questionnaire)
– 84 PV projects
– 9 countries
– Different business concepts
• Complemented with:
– Results from the Solar Bankability project [1]
– General & public domain industry knowledge
[1] The Solar Bankability project is funded by the European Union’s Horizon 2020 research and innovation
programme under the grant agreement No 649997.

Financial Models
Energy flow from sunlight to consumer of electrical power:
Several sub-models
& assumptions
=
Uncertainties

Financial Models
PV plant design
• Solar
Resource
• P50/P90
• Degradation
behavior
• Availability
Procurement
process
• Technical
specifications
• Specific
requirements
• Factory
inspections
• Product
testing
Plant
construction
• Transportation
& handling
• Storage of
components
on-site
• Construction
supervision
Plant
acceptance
• Protocol for
visual
inspection
• Relevant
equipment (IR
/ EL)
• Provisional
acceptance
• Protocol for
data collection
and KPIs
calculation
Operations &
Maintenance
• Monitoring
capabilities
• Preventive &
corrective
maintenance
protocols
• Relevant
equipment (IR
/ EL)
Weaknesses when dealing with technical assumptions and risks in PV
financial models:

Weaknesses: PV Plant Design
Solar Resource Assessment
Effect of long-term trends in the solar resource is often not fully
accounted for
Forecast of future long-term irradiation based on the average of 32 meteo stations in the Netherlands

Uncertainties calculation
Exceedance probabilities (e.g. P90) are often calculated assuming a
normal distribution for all elements contributing to the overall
uncertainty
Cumulative distribution function for the long-term (58 years) GHI average of 32 meteo stations in NL

• Incorrect degradation rate and inaccurate rendering of the system
behavior over time
• Incorrect availability assumptions used to calculate the initial
yield for the project investment financial model as opposed to the
O&M plant availability guarantee

Recommendations: PV Plant Design
• Project Pre-Feasibility
– Re-run financial models as the plant takes physical shape on the design
table
• Plant Design
– Quality control is critical for enabling realization of financial plan
– Finding and correcting errors at design phase are inexpensive (x 10 rule)
– Ensure back to back guarantees between EPC and O&M
– Minimize events uncovered by guarantees

Weaknesses: Procurement Process
• Technical specification of the PV plant components usually
consists only of a high-level description
• Project specific requirements such as salt-mist, ammonia or
resistance to PID, with the relevant IEC certification testing, are not
always specified
• Lack of specifications requiring factory inspection or product
testing

Weaknesses: Plant Construction
• Absence of standardized transportation & handling protocol
• Inadequate quality procedures in component un-packaging and
handling during construction
• Lack of construction supervision

Weaknesses: Plant Acceptance
• Inadequate protocol for data collection, KPIs calculation and
visual inspection
• Lack of relevant equipment (IR / EL)
• No short-term performance test at provisional acceptance

Recommendations
• Procurement & Construction
– Ensure quality control for the hardware, the workmanship, and software
– Apply quality control before leaving manufacturing factory and upon site arrival
• Acceptance
– EPC warranty should ideally be 24-months to enable discovery of deficiencies
during the first year and the effect of their correction during the 2nd year
– Deficiencies found during acceptance may cost:
• 10 times more than during construction
• 1/10 of what would cost to rectify during O&M

Weaknesses: Operations
• Selected monitoring system is not capable of advanced fault
detection and identification
• Missing or inadequate maintenance of the monitoring system
• Preventive & corrective maintenance protocols
• Inadequate or absent relevant equipment (IR / EL)

Recommendations: Operations
• Operations
– Ensure advanced monitoring system and O&M operators ,who act on the
monitoring system's signals
– Without accurate and advanced monitoring, there is little possibility for
optimizing operational activities
– First years of operation are critical for meeting the financial model:
• 1st year for finding faults and repairing them
• 2nd year for verifying that the faults have been rectified
• 3rd year of contractual obligation to enable a smooth transition from
warranty period to the “bottom of the bathtub” curve.

Long-Term Yield Estimates & Their
Level of Confidence
Initial P50 and P90 yield estimates vs actual electricity production for a
portfolio of 41 PV plants
Outliers ?

Level of Confidence
• Deviations below the confidence margin (P90) disappear after
correction
• 1.13% over-estimation (or under-performance?)

Level of Confidence
• Results at plant level:
– For most of the PV plants across the analyzed portfolio the actual
electricity production during first year of operation lies within expected
uncertainty ranges
– Actual availability is lower than initial estimates in LTYA

Level of Confidence
• Results at portfolio level:
– The yield is slightly lower than initially estimated during the design
phase (-1.15%)
– The dispersion (nRMSE) is around 4.4% for the analyzed portfolio
– These deviations are typically expected to be mainly due to the
variability of the solar resource and other site specific losses that
are not precisely modelled during the design phase

Risk Quantification
Risk
OPEX
high
low
Risk
lowlow
CAPEX
OPEX
high high high
low lowlow
Systematical approach required

Risk Quantification
• Long-Term Yield Estimate with level of confidence
• Feasibility study based on project-specific assumptions 1
• Cost-based FMEA based on historical and statistical data2
– Cost Priority Number (CPN)
1 Mitigating Financial Risks in a PV Investment, Mike Green (next
presentation)
2 www.solarbankability.org

Risk Quantification by CPN
• First approach to implement a cost-based FMEA to the
PV sector
• Based on statistical analysis and assumptions
• Cost of loss from system downtime (Costloss)
• Cost for detection and mitigation (Costfix)
• Further factors: occurrence, irradiance, power loss, PPA,
etc…
 Tool to benchmark mitigation measures:
www.solarbankability.org

Conclusions
• Main areas of weakness in a PV project:
– Inaccurate assumptions for the future behavior of the PV
plant
– Lack of methods to enable these assumptions during
design, building, commissioning, and O&M
– Lack of quality control throughout the plant’s life
• The benefit of individual mitigation measures can be
calculated using e.g. CPN method 1
1 www.solarbankability.org

Conclusions
• Keys to ensure that the financial model holds true:
– Accurate assumptions for the future behaviour of the PV
plant
– Ensure these assumptions are enabled during design,
building, commissioning, and O&M
• Need of high-level of quality control throughout the plant’s
lifetime
• Focus on the technical aspects of the EPC and O&M scopes of
work to manage the technical risks linked to the CAPEX
and OPEX of PV investments.

Find more results in Task 13 Report
TECHNICAL ASSUMPTIONS USED IN PV FINANCIAL MODELS
REVIEW OF CURRENT PRACTICES AND RECOMMENDATIONS
Download the report:
http://www.iea-pvps.org/index.php?id=426

Task 13 Workshop at EU PVSEC 2017
Save the date!
Learn more about our work during the IEA PVPS Task 13 Workshop at
EU PVSEC 2017 in Amsterdam, The Netherlands
PV System Performance and PV Module Reliability
Day: Tuesday, 26th September 2017
Time: 08:30 – 12:30
Site: RAI Convention & Exhibition Centre, Amsterdam
Access:Open to all registered participants

Thank you
Presented by:
Mauricio Richter
MRI@3e.eu

Mitigating Financial Risks in a PV Investment
Mauricio Richter (3E), Jan Vedde (SiCon)
TASK 13: PERFORMANCE AND RELIABILITY OF PV SYSTEMS
Download the report:

Financial Risks to a PV Investment
A PV project can be divided into 2 distinct stages:
1. FeasibilityDesignLicenseFinance
Shovel Ready
2. BiddingBuildingCommisioningMaintaining
PV project risk can be divided into 2 distinct causes:
• Uncertainties in the design data
• Lack of Quality Control

Feasibility – based on parameters:

Short list of some PRODUCTION parameters:
System Parameters
1. Module power
2. Solar Resource
3. Ground/roof coverage
4. Module tilt
5. Array azimuth
Loss Parameters:
1. Shading
2. Soiling
3. Temperature
4. Initial degradation
5. Annual degradation

Short list of some PROJECT parameters:
Project Parameters
1. Licencing costs
2. Installation costs
3. Unavailability
4. O&M fixed costs
5. O&M contingencies

Primary Equations:
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 = 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 − � 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = 𝒇𝒇(𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸, 𝑇𝑇𝑇𝑇 𝑇𝑇𝑇𝑇, 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶, 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡)
 What is the real Solar Resource over 20+ years?
 What are the accurate values for each of the losses?

Garbage In Garbage Out=
An energy simulation
is only as good as the
input data

Most parameters used in simulations are accompanied by a variance:

Uniform Distribution -0/+3%

P90
P50
Normal Distribution

Triangular Distribution

P90
P50
Triangular Distribution

0,0%
0,3%
0,6%
0,9%
1,2%
1,5%
1,8%
2,1%
2,4%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Cumulativeprobability
270 271 272 273 274 275 276 277 278 279
Module Power [Wp]
Pareto Plot for Module Power [Wp]
10,00000%(x=270,81)
50,00000%(x=274,03)
0%
1%
2%
3%
4%
5%
6%
7%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
850 900 950 1.000 1.050 1.100 1.150 1.200 1.250
Horizontal global irradiation [kWh/m2/year]
Pareto Plot for Horizontal global irradiation [kWh/m2/year]
10,00000%(x=984,49)
90,00000%(x=1116,28)
10,00000% 80,00000% 10,00000%
0%
1%
2%
3%
4%
5%
6%
7%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0,138 0,141 0,144 0,147 0,150 0,153 0,156 0,159 0,162
Irradiation increase due to tilt angle [%]
Pareto Plot for Irradiation increase due to tilt angle [%]
10,00000%(x=0,15)
90,00000%(x=0,15)
10,00000% 80,00000% 10,00000%
0,0%
0,5%
1,0%
1,5%
2,0%
2,5%
3,0%
3,5%
4,0%
4,5%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
30% 32% 34% 36% 38% 40% 42% 44% 46% 48% 50%
Ground Coverage Ratio [%]
Pareto Plot for Ground Coverage Ratio [%]
10,00000%(x=34,71%)
90,00000%(x=45,68%)
10,00000% 80,00000% 10,00000%
a
)
b
)
c
)
d
)
0%
1%
2%
3%
4%
5%
6%
7%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1,86% 1,89% 1,92% 1,95% 1,98% 2,01% 2,04% 2,07% 2,10% 2,13%
Irradiation loss due to shading [%]
Pareto Plot for Irradiation loss due to shading [%]
10,00000%(x=1,96%)
90,00000%(x=2,04%)
10,00000% 80,00000% 10,00000%
0,0%
0,3%
0,6%
0,9%
1,2%
1,5%
1,8%
2,1%
2,4%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0,0% 0,3% 0,6% 0,9% 1,2% 1,5% 1,8% 2,1% 2,4% 2,7% 3,0%
Irradiation loss due to soiling [%]
Pareto Plot for Irradiation loss due to soiling [%]
10,00000%(x=0,31%)
90,00000%(x=2,70%)
0%
1%
2%
3%
4%
5%
6%
7%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1,44% 1,46% 1,48% 1,50% 1,52% 1,54% 1,56% 1,58%
Module loss due to temperature [%]
Pareto Plot for Module loss due to temperature [%]
10,00000%(x=1,48%)
90,00000%(x=1,52%)
10,00000% 80,00000% 10,00000%
0,0%
0,5%
1,0%
1,5%
2,0%
2,5%
3,0%
3,5%
4,0%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0,0% 0,5% 1,0% 1,5% 2,0% 2,5% 3,0% 3,5% 4,0% 4,5% 5,0%
Module loss due to LID/deviation from nominal power [%]
Pareto Plot for Module loss due to LID/deviation from nominal power [%]
10,00000%(x=0,99%)
90,00000%(x=3,78%)
10,00000% 80,00000% 10,00000%
a
)
c
)
d
)
0,0%
0,3%
0,6%
0,9%
1,2%
1,5%
1,8%
2,1%
2,4%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
600 620 640 660 680 700 720 740 760
Turn-key installation cost in total [EUR/kWp]
Pareto Plot for Turn-key installation cost in total [EUR/kWp]
10,00000%(x=614,97)
90,00000%(x=734,66)
0,0%
0,5%
1,0%
1,5%
2,0%
2,5%
3,0%
3,5%
4,0%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0,0% 0,3% 0,6% 0,9% 1,2% 1,5% 1,8% 2,1% 2,4% 2,7% 3,0%
Technical unavailability [%]
Pareto Plot for Technical unavailability [%]
10,00000%(x=0,54%)
90,00000%(x=2,22%)
10,00000% 80,00000% 10,00000%
0,0%
0,5%
1,0%
1,5%
2,0%
2,5%
3,0%
3,5%
4,0%
4,5%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0,0% 0,1% 0,2% 0,3% 0,4% 0,5% 0,6% 0,7% 0,8%
PV system (module) degradation [%/yr]
Pareto Plot for PV system (module) degradation (%/yr)
10,00000%(x=0,20%)
90,00000%(x=0,64%)
10,00000% 80,00000% 10,00000%
0%
1%
2%
3%
4%
5%
6%
7%
Probability
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
30.000 32.000 34.000 36.000 38.000 40.000 42.000 44.000 46.000 48.000 50.000
O&M - fixed yearly fee [EUR/year]
Pareto Plot for O&M - fixed yearly fee
10,00000%(x=37480,90)
90,00000%(x=42566,23)
10,00000% 80,00000% 10,00000%
a
)
b
)
c
)
d
)

A Feasible Outcome
First Year Energy Production Revenue from Energy Sales
Project IRR by CAPEX Leveraged IRR after Tax

A Feasible Outcome
First Year Energy Production
P90
P50

ing operation (conversion efficiency) [%]
Irradiation loss due to shading [%]
Turn-key installation cost in total [EUR/kWp]
B1. O&M - fixed yearly fee [EUR/year]
Ground Coverage Ratio [%]
Module loss due to temperature [%]
Irradiation increase due to tilt angle [%]
Technical unavailability [%]
Module Power [Wp]
ue to LID/deviation from nominal power [%]
PV system (module) degradation (%/yr)
250 260 270 280 290 300 310 320
Conditional Mean. (scale factor = thousands)
Tornado Plot for Power production delivered to the grid [MWh]
Visual Aids

Module loss due to LID/deviation from nominal power [%]
PV system (module) degradation (%/yr)
12,9
13,2
13,5
13,8
14,1
14,4
14,7
15,0
15,3
15,6
ConditionalMean.scalefactor=millions
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Spider Plot for Revenue from power sales [EUR]
Visual Aids
Mean

Shovel Ready
PV project risk can be divided into 2 distinct causes:
 Uncertainties in the design data
• Lack of Quality Control



Quality Control
The Anderson Rule of 10:

 Strategize
 Classify
 Understand
 Manage
Mitigating the Financial Risks - QC

Strategize:
 Top level decision makers MUST take ownership of the QC
management for ensuring CONTINUETY from the
BEGGINING
 Define what must be done to ensure
 Continuity
 Desired level of product
 Assign the resources to achieve these goals

Classify:
 Ensure a wide variety of skill sets in the design team
 Brainstorm
 Checklists to ensure all potential risks are identified
 Classify risks in terms of occurrence frequency and severity
in terms of financial impact

Understand:
 Analyze the root cause(s) of the various risk factors and
interrelations between them
 Identify the most important influencers that may challenge
the financial performance of the project.
Manage!

 Analyzed the technical assumptions used in a financial
model
 We have analyzed the areas of weakness in a PV project
 We have seen some levels of accuracy between design
calculations and finished project
 Quantified the cost of a technical risk
 We have seen new methodology for calculating the benefit of a
mitigating measure using cost-based FMEA
 Use of CPN for evaluating the cost of a mitigating action
Conclusions:

 Mitigating financial risk in the business plan
 Challenge the operators of the simulation program used for
calculating first year energy generation
 Calculate the major indices of the business plan using both the
parameter and its variance
 Use graphing tools to visualize the interdependence of these
variables
 Ensure that a rigid QC regimen exists that will enable the design
parameters on which the business plan was built remain intact
from feasibility through to O&M
Conclusions:

IEA-PVPS Task 13

Technical Assumptions Used in PV Financial Models: Review of Current Practices and Recommendations

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Technical Assumptions Used in PV Financial Models: Review of Current Practices and Recommendations

Similar to Technical Assumptions Used in PV Financial Models: Review of Current Practices and Recommendations (20)

More from Leonardo ENERGY

More from Leonardo ENERGY (20)

Recently uploaded

Recently uploaded (20)

Technical Assumptions Used in PV Financial Models: Review of Current Practices and Recommendations