A review of DECC s Impact Assessment of Feed-in-Tariff rates for small-scale renewables

A review of DECC s Impact Assessment of Feed-in-Tariff rates for small-scale renewables A report for the British Photovoltaic Association 22 October 2015 Cambridge Econometrics Covent Garden Cambridge CB1 2HT UK 01223 533100 alb@camecon.com www.camecon.com

Cambridge Econometrics mission is to provide rigorous, accessible and relevant independent economic analysis to support strategic planners and policy-makers in business and government, doing work that we are interested in and can be proud of. Cambridge Econometrics Limited is owned by a charitable body, the Cambridge Trust for New Thinking in Economics. www.neweconomicthinking.org 2

Authorisation and Version History Version Date Authorised for release by 2.0 22/10/15 Philip Summerton Final Report. Description 3

Contents Page Executive Summary 5 1 Scope 7 2 Review of Evidence 8 3 Alternative options 16 Appendix A Data table 21 4

Executive Summary Higher Bands The British Photovoltaic Association (BPVA) commissioned Cambridge Econometrics (CE) to review the data and assumptions applied in DECC s Impact Assessment of proposed revisions to the Feed-in-Tariffs for small scale renewables (2015). The review focussed on small scale (domestic) solar PV. The analysis suggests that some of the assumptions and data used in the Impact Assessment of solar PV are either not suitable or not robust: As a result of proposed changes to the banding classifications, the OPEX for a representative household 3kWp system are significantly underestimated. An electricity price has been applied to calculate the bill saving that includes not only the variable unit cost, but also the fixed standing charge. The calculations do not include the impact of long term system degradation. The load factor used for the impact assessment calculation was for a household with the optimal possible set of characteristics in the UK; that of a household based on the south coast with ideal conditions for solar panels. The targeted rate of a return for this well-sited installation on the south coast was given at a 4% real return per annum compared to 8% in the 2012 Impact Assessment of Feed-in-Tariffs. The Parsons Brinkerhoff report upon which the estimate was based reports a central estimate of 6.2% hurdle rate for a potential domestic installer. The implications for the Feed-in-Tariff of the assumptions selected was significant: The result of the DECC impact assessment was for a Feed-in Tariff rate of 1.64p/kWh for small scale solar PV. Our calculations, which are based on the alternative assumptions suggested in this report, show that this would provide an annual real rate of return of just 1.5% for the best sited locations in the country, and a negligible 0.4% for the average household. In order to provide potential investors with a rate of return in line with the previous literature, the FIT rates would need to be increased above those proposed in the impact assessment. A FIT rate of 7.92p/kWh would provide a 4.8% return to the average investor, and a 6.2% return to the best sites, whereas a FIT rate of 9.11p/kWh would provide a 6.2% return to the average investor, and a 7.1% return to the best sites under our revised assumptions. Similar analysis was also carried out for the higher tariff bands, with the minimum FiT rates required to give a 4% rate of return calculated. A rate of 6.06p/kWh is required in the 4-50kWp band, 4.43p/kWh rate is required in the 50-250kWp and 250-1000kWp bands, and 2.80p/kWh for the 1000kWp+ band. These calculations were based on values taken at face value from DECC Impact Assessment, and may therefore be subject to revision. 5

Degression Rates and Expenditure Projections Grants for electricity storage Two scenarios were analysed for the degression projections based on a starting 7.92 FiT rate. Scenario A featured a 10% quarterly regression and scenario B featured a 5% quarterly regression. These projections were then used to calculate expected total expenditure under different take-up assumptions. Expenditure projections for DECC s own proposed tariff rates and current tariff rates are also presented for comparison purposes. It was found that both scenario A and B reduced projected expenditure from current levels by over 40%. The BPVA has proposed that finance is used from the DECC innovation support scheme for energy storage towards supporting the installation of energy storage systems in the form of a one off grant payment, applicable only to new installations and when used in combination with a new PV installation. This would be an alternative method of providing a more realistic rate of return to potential investors, however significantly improved FiT rates would still be necessary. 6

1 Scope 1.1 Scope Cambridge Econometrics (CE) was commissioned by the British Photovoltaic Association (BPVA) to review the data and modelling approach applied by DECC in its Impact Assessment 1 of feed-in-tariff options, and in the underlying work carried out by Parsons Brinkerhoff in the publication Small-scale Generation Cost Update 2. In particular, the BPVA was particularly concerned about the choice of input assumptions that resulted in a proposal to reduce the Feed-in-Tariff for smallscale domestic (0-10kWp) to 1.63p/kWh. After an initial assessment of the data and the modelling approach, it was agreed that the review would focus on five aspects: degradation OPEX and size banding electricity prices load factors hurdle rates (real internal rates of return) A second stage of the analysis assesses the implications of alternative proposals for feed-in-tariffs on internal real rates of return for households in different parts of the country. We also consider the implications of the different options for take-up of solar PV and the number of installations that can be supported under different options for caps. 1.2 Layout The layout of this report follows the scope: Chapter 2 presents our review of the assumptions and calculations used in the Impact Assessment, and Chapter 3 sets out our assessment of the alternative options put forward by the BPVA. 1 https://econsultation.decc.gov.uk/office-for-renewable-energy-deployment-ored/fit-review- 2015/supporting_documents/IA%20for%20FITs%20consultation%20August%202015%20%20FINAL%20do cx%20esignature%20included%20corrected.pdf 2 https://econsultation.decc.gov.uk/office-for-renewable-energy-deployment-ored/fit-review- 2015/supporting_documents/SmallScale%20Generation%20Costs%20Update.PDF 7

2 Review of Evidence It is noted that the CAPEX costs, which represents the most significant part of the system cost, presented in the Impact Assessment for small scale (<10kWp) solar PV are broadly reasonable. However, many of the supporting assumptions used in the calculation of the revised Feed-in-Tariffs are reviewed and critiqued below. 2.1 Degradation Degradation of the performance of the PV system is not considered DECC does not address the issue of solar panel degradation in the Impact Assessment, despite assuming a long lifespan of 30 years, and neither is it addressed in the Parsons Brinkerhoff study. Standard estimates put degradation at 1% during the first year, followed by around 0.5% degradation thereafter. Indeed, DECC acknowledge the importance in modelling panel degradation in their 2012 Renewables Obligation banding review 3 : The degradation of solar PV panels over time was not something that was included within our original modelling. The load factor assumptions for both large-scale ground-mounted and building-mounted solar PV now include an annual 0.5% rate of panel degradation which is consistent with the warranties available from manufacturers. Accounting for degradation has a significant effect on the modelling of total costs over the full lifetime of the PV system. After including degradation in our model, we found that the present value of the total revenue over the lifetime of the system is reduced by 348 and this implies that the feed-in-tariff required to meet DECC s target IRR would be an additional 0.86p/kWh (see Table 2.1). Table 2.1. Effect of including degradation rate Required FiT to give Rate of degradation Total revenue over 4% real internal rate pa lifetime of system of return DECC IA assumption 0% 5548 1.64p/kWh Revised assumption 0.5% 5204 2.50p/kWh 2.2 Operating costs (OPEX) and banding An incorrect value for OPEX cost is used In DECC s Impact Assessment it is asserted that the OPEX costs used in the DECC FIT calculation model are based on the estimates presented in the Parsons Brinckerhoff independent report. Based on five survey responses and a literature review, PB assumes that domestic OPEX is the cost of replacing the inverter, which PB assumes need replacing every 10 years. PB assumes inverter costs of 1000 for the 0-4kWp band, and 1200 for the 4-10kWp band. All system cost estimates are based on average system size of 3.06kWp for the 0-4kWp band and 8.21kWp for the 4-10kWp band (see Table 2.2). 3 https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/66516/7328-renewablesobligation-banding-review-for-the-perio.pdf 8

Table 2.2. OPEX calculation for different bands Band Inverter cost Average system OPEX/kWp/year size 0-4kWp 1000 3.06 (1000 10) 3.06 = 32.7 4-10kWp 1200 8.21 (1200 10) 8.21 = 14.6 The OPEX estimates from the PB report are also reported in DECC s impact assessment (inflated to 2016 prices), with low and high cases calculated using a 20% difference each way: Table 2.3. OPEX estimates at different bands DECC has proposed in its Impact Assessment to merge the banding for <4kWp systems and 4-10kWp systems to a single category of 0-10kWp systems. One implication of doing this is that the fixed OPEX for replacing an inverter is proportionately less of the total cost of the larger (6kWp) average system that represents the new banding category - for which the applied assumption of 20/kWp/year is valid. However, it is not a valid assumption when applied to a 3kWp system which is the typical household installation and the OPEX should have been recalculated in line with the stated methodology presented by DECC in the IA Where tariff bands have changed, the underpinning assumptions are broken down by installation size and have then been re-calculated from the raw data gathered by PB. The result is that 3kWp systems would not be a viable financial proposition because the OPEX cost, which reflects the cost of a replacement inverter, is much higher than estimated in the DECC calculation of the proposed FIT. Had the OPEX been recalculated correctly for the 3kWp system, PB s method would give a cost estimate of 102/year (or 34/kWp/year as stated, but not applied, in the Impact Assessment) for a 3kWp system. The difference in the present value (discounted at 4% over 30 years) of the corrected total annual OPEX estimation is 501. This would require the Feed-in-Tariff to be revised to 2.89p/kWh, an increase of 1.25p/kWh (see table 2.4). Table 2.4. Effect of including alternate OPEX assumptions Annual OPEX Total Opex Cost over lifetime of system Required FiT to give 4% return DECC IA Assumption 73 1260 1.64p/kWh Corrected assumption 102 1764 2.89p/kWh 9

2.3 Electricity prices DECC applies an electricity bill saving which includes an implicit saving in fixed standard charges DECC overestimates the bill savings to the householder by factoring the total price of electricity in the Impact Assessment analysis. Instead of using the (average) variable unit cost of electricity to calculate the total bill savings, DECC applies the average cost (including both standing charges and variable unit costs). DECC also reports the following in the Impact Assessment: Bill savings are valued using the central retail electricity prices in the Green Book supplementary guidance. These are set out in table 13 below. We have taken an average of the services and industrial prices for the relevant bands in the absence of information about the sector of the installations in those bands. More information about which installation group faces which electricity price is set out in paragraph 5.28 below. The medium electricity prices have been used within this analysis. These are due to be revised and republished by DECC during the period of the consultation. This will be taken into account in the analysis for the Government response to the consultation. Since the publication of DECC s retail prices, declines in fossil fuel price have led to falls in the electricity wholesale price. Other things being equal, this would result in a decrease in the bill savings for a FITs installation and therefore an increase in the generation tariff to reach a particular point on the supply curve for FITs installations. While in the short term electricity prices may be lower, this may not be true over the longer term. Given that FITs installations may continue to generate for up to 35 years, DECC has decided to continue using central electricity price projections for this analysis. Table 2.5. Electricity Prices used in DECC Impact Assessment (p/kwh, 2016 prices) Sector 2014 2015 2016 2017 2018 2019 2020 Wholesale 4.8 5.5 5.7 5.4 5.2 5.3 5.7 Residential 16.6 16.3 17.4 18.4 18.6 19.8 19.8 Service/ Industrial 9.1 10.1 10.9 10.9 11.0 12.3 12.4 However the data shown in Table 2.5 represents average electricity costs, not average variable electricity costs. Hence it is not the correct data to use in such an assessment because the standing charge will not be reduced in line with the reduction in electricity demand. The use of solar panels to generate electricity for domestic consumption will not affect the householder s obligation to pay a standing charge, therefore only the variable unit price should be included in an analysis of any cost savings that are made. A table of average variable unit prices across the UK was produced by DECC in 2014 and is shown below. The average variable unit price for the UK is given as 0.14 /kwh. This therefore, is the most appropriate number to use when calculating the expected revenues from domestic solar panel installation. 10

Table 2.6. Average variable unit costs and standing charges for standard electricity in 2014 4 Region Average variable unit price ( /kwh) (5) East Midlands 0.13 Eastern 0.13 London 0.14 Merseyside & North Wales 0.15 North East 0.14 North Scotland 0.15 North West 0.14 Northern Ireland 0.17 South East 0.14 South Scotland 0.13 South Wales 0.15 South West 0.15 Southern 0.14 West Midlands 0.14 Yorkshire 0.14 United Kingdom 0.14 Using the variable unit cost of 14p/kWh in the impact calculation instead of 16p/kWh price leads to a significant difference in the resultant FIT rate required to provide a real rate of return of 4% pa, even assuming the same above inflation real electricity price rises from that point on. The reduction in the amount of bill savings given by using a more accurate estimate of electricity prices is 528 and the corresponding increase in FiT required to give a rate of return of 4% is 1.31p/kWh. Table 2.7. Effect of alternate electricity price assumptions Variable unit electricity Price Bill Savings over lifetime of system Required FiT to give 4% return DECC IA Assumption 16p/kWh 4,210 1.64p/kWh Alternative Assumption 14p/kWh 3,699 2.95p/kWh 4 https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/416987/table_224.xls 11

2.4 Load factors The load factor assumption has a significant impact on the calculation of the return of a solar PV investment Load factors affect the FIT calculation in that an over-estimation of the load factor leads to a significant underestimation in the size of the FIT required to provide a realistically attractive rate of return. The DECC impact assessment does not use the central case assumption of 10.3% for the Midlands as previous studies and the previous FIT rate Impact Assessment (2012) have done. Instead, it uses the 11.3% high scenario for the South West (Exeter). The proposed justification for this is the intention of targeting well-sited installations. Regarding the choice of load factor, DECC reports the following: The load factor is the proportion of time for which an installation is expecting to generate electricity. The 2015 PB analysis shows that average load factors have increased for solar PV, Hydro, Wind and AD, compared to assumptions from the 2012 Review.18 It is assumed that the load factors are constant over time. Table 2.8: PV Load factors (DECC Impact Assessment Table 5) 2015 2012 Low Central High Central Scotland 8.4% 8.9% 9.5% 9.0% Midlands 9.7% 10.3% 10.9% 9.5% South East 9.9% 10.5% 11.1% 9.7% South West 10.1% 10.7% 11.3% 10.0% Load factors are taken from the higher range of PB data. This reflects the intention of targeting well-sited installations. For PV, the 11.3% estimate is representative of installations located in the South West with high load factors. The load factor estimates are based on the report from Parsons Brinkerhoff (PB). However PB issue a warning about the quality and validity of their data: The load factor analysis above is based on a limited number of data points and on PVGIS irradiance data which has limitations. Both of these variables are applied at a regional level and as such, load factors at specific project locations may vary. Parsons Brinckerhoff acknowledges that load factors for PV installations could be refined further using other analytical methodologies should additional robust data be sourced. The difference in FiT rate required to accommodate the central case in the midlands instead of only the most optimally sited areas of the South West is significant: an extra 1.50p/kWh (see Table 2.9) Table 2.9. Effect of alternative load factor assumption on required FiT rate Load Factor Electricity Generated pa Required FiT to give 4% return DECC IA Assumption 11.3% 2972kWh 1.64p/kWh Alternative assumption 10.3% 2709kWh 3.14p/kWh 12

2.5 Hurdle rates DECC sets out the hurdle rates for domestic solar PV (0-10kWp) in Table 7 of the Impact Assessment on p13 (see Table 2.10 below). Table 2.10: Hurdle rates for domestic investors (DECC Impact Assessment, Table 7) 2015 2012 Minimum 2.5% 3.5% Average 6.2% 8.0% Maximum 10.0% 12.5% The hurdle rates chosen are below the ones recommended in the consultation report These rates are sourced from the analysis by Parsons Brinkerhoff and Ricardo-AEA. That study states that the domestic hurdle rates were gathered from only seven survey responses from households considering or having recently installed solar PV. The consultants did not receive data from commercial installers to derive the commercial hurdle rates and it is not particularly clear what this data is based on. Despite presenting these hurdle rates, DECC has used a domestic hurdle rate of 4% rather than the central rate of 6.2% presented in the supporting analysis. In the Impact Assessment DECC argues the following The target rates of return have been set as the maximum of the low end of hurdle rate ranges for domestic and commercial investors (please refer to table 7 in section 4 for a reminder of the full range of hurdle rates). For example the low end of the hurdle rate ranges for domestic and commercial investors in PV are respectively 2.5% and 4%. The maximum of these two values (4%) is therefore identified as the target rate of return for solar PV. The focus on domestic and commercial hurdle rates reflects the scheme s objective to support primarily non-energy professionals (as per our State aid approval); and the choice of level corresponds to targeting a level that is high enough to allow deployment to come forward, but low enough to avoid overcompensation. In summary, DECC has selected a hurdle rate for the domestic sector which is based on no survey responses, is for the commercial sector and is not fully explained in the supporting documentation. The supporting evidence report by PB notes Suitable hurdle rates for commercial and developer/utility-scale installations were subsequently calculated based on the assumption that commercial investors may also invest for reasons apart from financial return, although rates higher than domestic investors are proposed. Developers will focus on returns, but with competitively priced finance (e.g. debt for 70% of the project at 6% interest and equity returns at 8.5%) real hurdle rates as low as 4% could arise. In other words, the supporting evidence suggests that if investors have other incentives or reasons for installing solar PV that are not financial it is possible to assume a very low real hurdle rate. The selection of a minimum rate will severely limit take-up because only perspective installers that have a return lower than (or equal to) the minimum 13

rate will go ahead with the project. The analysis suggests that the minimum and maximum are defined as the range falling within one standard deviation of the mean after removing outliers, but as there were only seven household respondents and no commercial respondents (upon which the selected rate is based), the ability to justify the reduction in hurdle rates from 2012 based on the sample size presented seems extremely limited. Moreover, the 2012 Impact Assessment on Feed-in-Tariffs suggested a real hurdle rate of 8% for small scale domestic installations, with 95% of perspective investors requiring a rate between 3.5% and 12.5%. This has the implication that only 2.5% of perspective investors would consider PV if the real return on investment was less than or equal to 3.5%. Given the lack of data, it would be more appropriate to use the average (mean) real hurdle rate of the domestic survey respondents (6.2%) given that the distribution cannot be statistically determined from such a small sample and that no survey response were received for commercial rates (on which the domestic estimates were based). Another alternative would be to determine the desired take up based on the distribution of hurdle rates presented in the 2012 Impact Assessment as it is unlikely that the required hurdle rate would substantially change over time. The selection of the hurdle rate has substantial consequences for the calculation of the feed-in-tariff that delivers an internal (real) rate of return equivalent to the real hurdle rate. The difference between DECC selecting a 4.0% real hurdle rate and the average domestic real hurdle rate of 6.2% adds 3.26p/kWh to the solar feed-in-tariff rate (see Table 2.11). Table 2.11 Effect of alternative hurdle rate on required FiT rates Real Hurdle Rate Required FiT to give return in line with real hurdle rate DECC IA assumption 4% 1.64p/kWh Alternative assumption 6.2% 4.90p/kWh 2.6 Summary The combined effects of the omission of a treatment of degradation, the incorrect treatment of electricity prices and Opex results in a significant difference in the real rates of return available at the proposed FiT rate of 1.63p/kWh. Figure 2.1 shows the necessary steps required to bridge the gap between the DECC proposal rate of 1.63p/kWh and proposed tariffs of 7.92p/kWh and 9.11p/kWh. Correcting for the three technical errors in the Impact assessment (detailed in sections 2.1 to 2.3 above), requires an increase in FiT rate to 5.05%. However this still only provides a rate of return of 4% in even the best located sites in the UK, and a significantly lower return for the majority of UK households. In order to raise the rate of return to the PB recommended required hurdle rate of 6.2%, an additional increase to 7.92p/kWh is required. To achieve this level of return for projects sited in the Midlands the rate would need to be increased to 9.11p/kWh. 14

Figure 2.1. Required Adjustments to correct Feed in Tariff 15

3 Alternative options 3.1 Analysis of alternative options The BPVA has suggested a number of alternative FiTs for both small scale domestic PV and commercial scale PV. Table 3.1 shows the resultant rates of return for different sites in the UK for three different FiT rates in the 0-4kWh band (which the BPVA recommend maintaining) for corrected assumptions on degradation, variable unit electricity prices and operating costs. Table 3.1: Rates of returns corresponding for different load factors and FiT rates for 0-4kWh band FiT Rate Rate of Return South West High Scenario Rate of Return Midlands Average Scenario Rate of Return Scotland Low Scenario 1.63p/kWh 1.5% 0.4% -1.8% 6.58p/kWh 5.2% 4.0% 1.4% 7.92p/kWh 6.2% 4.8% 2.2% 9.11p/kWh 7.1% 6.2% 3.0% Higher Bands In Table 3.2, we present the assumptions used in the DECC impact assessment that provide a 4% rate of return for the representative UK household (Midlands average scenario) for the intermediate bands (4-50kWp, 50-250kWp and 250-1000kWp), and for the higher 1000kWp+ band. These bands correspond to the preferred bands recommended by the BPVA, and as such, these are the bands we have focused on for this analysis. We assume that the stand alone band will be designated the same tariff as the 1000kWp+ band for competitive reasons as outlined in DECC s impact assessment. The only significant correction to these figures from the DECC impact assessment calculation has been the inclusion of the omitted 0.5% annual degradation rate. All other assumption have been taken from the DECC Impact Assessment at face value and CE has not assessed the accuracy and validity of these assumptions. Further research, including insight from industry sources, could highlight errors or inconsistencies in these figures. We use the reference size installation used in the Impact Assessment, along with the CAPEX and OPEX costs stated and the lower electricity price for services/industry. Using these assumptions, we find that a rate of 6.06p/kWh is required in the 4-50kWp band, whereas a 4.43p/kWh rate is required in the 50-250kWp and 250-1000kWp bands, and 2.80p/kWh for the 1000kWp+ band. 16

Table 3.2: Assumptions and resultant FiT rates to provide a 4% return for centrally representative UK location 4-50kWp 50-250kWp 250-1000kWp 1000kWp+ Reference Size 30kWp 140kWp 455kWp 2840kWh CAPEX 1700/kWp 1500/kWp 1500/kWp 1300/kWp OPEX 10/kWp 10/kWp 10/kWp 10/kWp Annual Degradation 0.5% 0.5% 0.5% 0.5% Electricity Prices 10p/kWh 10p/kWh 10p/kWh 10p/kWh Required FiT Rate 6.06p/kWh 4.43p/kWh 4.43p/kWh 2.80p/kWh Rate of Return at Midlands Average Scenario Rate of Return at South West High Scenario 4% 4% 4% 4% 5% 5% 5% 5% Degression Rates Feed in Tariffs are slowly reduced over time so that the same rate of return is available in each quarter both as an incentive for, and a response to, the gradual reduction of costs of PV systems. The appropriate degression rate depends therefore upon the anticipated rate of reduction of CAPEX and/or OPEX costs. Here we make no predictions as to the likely rate of reduction of CAPEX or OPEX, we simply show the degressed tariff rates that would provide the same rate of return for two different rates of CAPEX reduction. In scenario A, a 10% quarterly degression rate would correspond to a 4.85% quarterly (20.9% annual) reduction in CAPEX costs. As an example, this would require CAPEX costs of 936/kWp for a 0-4kWp system by 2019 Q1, if there were no other changes in the costs or performance of the technology. Table 3.3: Suggested tariff rates with 10% degression 2016 2016 2016 2016 2017 2017 2017 2017 2018 2018 2018 2018 2019 q1 q2 q3 q4 q1 q2 q3 q4 q1 q2 q3 q4 q1 0-4kWp 7.92 7.13 6.42 5.77 5.20 4.68 4.21 3.79 3.41 3.07 2.76 2.49 2.24 04-50kWp 6.06 5.45 4.91 4.42 3.98 3.58 3.22 2.90 2.61 2.35 2.11 1.90 1.71 250kWp 4.43 3.99 3.59 3.23 2.91 2.62 2.35 2.12 1.91 1.72 1.54 1.39 1.25 50-250- 1000kWp 4.43 3.99 3.59 3.23 2.91 2.62 2.35 2.12 1.91 1.72 1.54 1.39 1.25 1000kWp+ 2.80 2.52 2.27 2.04 1.84 1.65 1.49 1.34 1.21 1.08 0.98 0.88 0.79 Stand Alone 2.80 2.52 2.27 2.04 1.84 1.65 1.49 1.34 1.21 1.08 0.98 0.88 0.79 17

In scenario B, a 5% quarterly degression rate would correspond to a 2.42% quarterly (10.0% annual) reduction in CAPEX costs. As an example, this would require CAPEX costs of 1267/kWp for a 0-4kWp system by 2019 Q1, if there were no other changes in the costs or performance of the technology. Table 3.4: Suggested tariff rates with 5% degression 2016 2016 2016 2016 2017 2017 2017 2017 2018 2018 2018 2018 2019 q1 q2 q3 q4 q1 q2 q3 q4 q1 q2 q3 q4 q1 0-4kWp 7.92 7.52 7.15 6.79 6.45 6.13 5.82 5.53 5.25 4.99 4.74 4.50 4.28 04-50kWp 6.06 5.76 5.47 5.20 4.94 4.69 4.45 4.23 4.02 3.82 3.63 3.45 3.27 50-250kWp 250-1000kWp 4.43 4.21 4.00 3.80 3.61 3.43 3.26 3.09 2.94 2.79 2.65 2.52 2.39 4.43 4.21 4.00 3.80 3.61 3.43 3.26 3.09 2.94 2.79 2.65 2.52 2.39 1000kWp+ 2.80 2.66 2.53 2.40 2.28 2.17 2.06 1.96 1.86 1.76 1.68 1.59 1.51 Stand Alone 2.80 2.66 2.53 2.40 2.28 2.17 2.06 1.96 1.86 1.76 1.68 1.59 1.51 Cap analysis In Table 3.5 below, we present the total expenditure implications over the period January 2016 to March 31 st 2019, for a number of combinations of proposed tariff rates and caps. We compare three different sets of generation tariff levels, one for each row of Table 3.5. DECC proposals show the effects of using the proposed generation tariffs from the Impact Assessment. CE Scenario A refers to the suggested tariff levels shown above in Table 3.3 with 10% quarterly degression and CE Scenario B refers to the suggested tariff levels in Table 3.4, with 5% quarterly degression applied. DECC Do Nothing Scenario uses current generation tariff levels and degression rates. For each set of generation tariffs we use three quarterly deployment scenarios to calculate forecasted expenditure for 2016-2019. The first deployment scenario based on the caps proposed by DECC in the Impact Assessment. The second deployment scenario is based on the higher caps recently proposed by the Solar Trade Alliance as part of their response to the FiT Consultation process. The third deployment scenario adopts the average quarterly deployment in 2014/2015 across the 2016-2019 period. 18

Table 3.5: Total expenditure under different deployment assumptions and Feed-in-Tariff assumptions Total Take-up 2016- FiT rate and 2019 (MW) degression proposals: Deployment Assumptions DECC caps STA caps 2014/15 take-up rate extrapolated 600 2700 2200 Capped total expenditure (2016-19) DECC 2015 14m 42m 42m BPVA Scenario A 37m 131m 156m BPVA Scenario B 44m 167m 187m DECC Do Nothing Scenario 79m 302m 324m As can be seen, both BPVA tariff scenarios would reduce expenditure by over 40% in comparison to the do nothing scenario. However, this relies on the assumption that deployment rates are maintained, but this is likely to be heavily influenced by the FiT rate on offer, rather than independent of it. 3.2 Analysis of the BPVA proposal for support scheme for energy storage The BPVA have proposed that finance is used from the DECC innovation support scheme for energy storage towards supporting the installation of energy storage systems in the form of a one off grant payment, applicable only to new installations and when used in combination with a new PV installation. This proposal would appear to be a good fit for the specifications of the scheme. The potential benefit to the consumer of this scheme and the required FiT rates required to give a selection of hurdle rates are shown in Table 3.6 below. The calculation below assumes a storage system cost of 500/kWh, an average increase in self-provision ratio of 23 percentage points, and a lifetime of 17 years or longer. 5 6 5 SMA, photovoltaik 2015 geschäftsmodelle für private und gewerbliche anlagen. 6 Deutsche Bank, https://www.db.com/cr/en/docs/solar_report_full_length.pdf 19

Table 3.6: FiT rates required to give various hurdle rates for different storage funding options No storage system 50% discount on 2000 4kWh storage system 100% discount on 2000 4kWh storage system Cost to government per consumer 6.2% rate of return for Midlands average 4.8% rate of return for Midlands average 4.0% rate of return for Midlands average 0 1000 2000 9.11p/kWh 7.45p/kWh 5.82p/kWh 7.92p/kWh 6.97p/kWh 3.99p/kWh 6.58p/kWh 5.09p/kWh 2.38p/kWh 20

Appendix A Data table Table A.1 shows a comparison between a recreation of the calculation used in DECC s impact assessment to produce the proposed figure of 1.63p/kWh for the residential sector, and an alternative calculation in which we propose a number of improvements to the assumptions used in the analysis. Note that due to precision errors (differences in rounding) we calculate a FiT rate of 1.64 p/kwh. Code Variable CE re-creation More DECC Calculation of DECC Source appropriate (table 21, IA) calculation figure 1 Size (kwp) 3kWp DECC IA table 10 2 Lifetime 30yrs DECC IA table 11 Does not correspond to any 3 Annual OpEx 73 value in the 102 assessment 4 Capex Cost per Undisclosed 1650 kwp calculation 5 Export Payments 4.85p DECC IA, para 4.22 DECC table 13, residential price 6 Electricity Bill projections. Does 16 p/kwh Initial Rate not distinguish 14 p/kwh between fixed and variable cost Real Electric Price 7 Rises 0.7% Implied from 6. 8 Degradation Rate 0% No justification 0.5% pa Minimum Scenario 9 Real Hurdle Rate 4% Commercial 6.2% Developer 10 CPI 2% DECC IA para 4.27 12 Load Factor 11.3% High scenario, South West 10.3% DECC IA, table 12, 13 Export Fraction 0.53 all building mounted category Source DECC IA table 9, <4kWp category average scenario Average variable unit costs and fixed costs for electricity for selected towns and cities in the UK (QEP 2.2.4) 7 Taken from publically available industry estimates DECC IA table 7, average scenario, Domestic user Average scenario, Midlands 14 Initial Capex 4950 4950 Implied from 1, 4 4950 Implied from 1, 4 15 Total Opex 1260 1262 Implied from 1, 2 1374 Implied from 1, 2 16 kwh/kwp 991 Implied from 12 903 Implied from 12 17 kwh in the year 2980 kwh 2972 kwh Implied from 1, 16 2709 kwh Implied from 1, 16 7 https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/416987/table_224.xls 21

18 kwh exported 1575 kwh Implied from 17, 13 1436 kwh Implied from 17, 13 19 kwh used 1397 kwh Implied from 17, 13 1273 kwh Implied from 17, 13 20 Export Income pa 76 Implied from 18, 5 70 Implied from 18, 5 21 Bill Savings pa 223 Implied from 19, 6, 7 178 Implied from 19, 6, 7 22 Total Cost 6210 6212 Implied from 14, 15 6324 Implied from 14, 15 23 Total Revenues 5550 5548 Implied from 20, 21 3344 Implied from 20, 21 24 Short Fall 660 664 Implied from 22, 23 2980 Implied from 22, 23 25 Equivalent Annual Figure 50 49 Implied from 24 247 Implied from 24 26 FiT Rate Required 1.63p 1.64p Implied from 25, 17 9.11p Implied from 25, 17 22