Responses to Budget & Finance Committee Questions on SEIP

Responses to Budget & Finance Committee Questions on the Pilot Solar Energy Incentive Program GoSolarSF Power Enterprise, SFPUC

Introduction to the Program

The San Francisco Public Utilities Commission proposed to initiate, as a pilot program, the Solar Energy Incentive Payment program, pending before Budget and Finance Committee. The pilot program would provide a financial incentive for San Francisco residents and businesses to install solar (photovoltaic) energy systems on their properties.

The pilot Solar Energy Incentive Program, announced in December 2007, and supported by the Public Utilities Commission in Resolution 08-0004, would provide from $3,000 to $5,000 for residential installations and up to $10,000 for commercial photovoltaic installations.

The pilot program would offer a one-time incentive payment for local solar electric projects to reduce the cost of installation.

The pilot program would have three distinct incentives for residential installations. The incentives are limits on assistance, meaning that the higher incentives are not additive to the lower ones, but are total incentives. The incentive for businesses is a simple capacity-based incentive.

Residents Businesses Basic Incentive $3,000 City installer incentive $4,000 $1.50 per Watt a system is designed to generate, up to a cap of $10,000. Environmental justice $5,000 incentive

Incentive payments are tied to individual electric meters, meaning that buildings with more than one meter or applicants owning more than one property are eligible for more than one incentive payment, subject to the provisions below. Any system whose California Solar Initiative incentive reservation date is on or after December 11th 2007 would be eligible for the pilot San Francisco incentive.

The basic incentive would be available to any resident installing solar generation on his/her own property located within San Francisco. One incentive would be allowed per meter.

The city installer incentive would exist to encourage the use of local installation businesses. It would be available to residents eligible for the basic incentive who contract with an installer holding a current San Francisco business registration certificate showing a San Francisco address as the place of business.

The environmental justice incentive would be available to installations on property in lower income neighborhoods considered “environmental justice districts” because of their proximity to industrial sites and major highways. The SFPUC will apply the environmental justice incentive payments to installations on properties located within the zip codes 94107 and 94124. 2

The SFPUC intends to fund the pilot program by redirecting $3 million allocated to the Mayor’s Energy Conservation Account toward this program. The Board of Supervisors were asked to endorse the program, and express their intent to appropriate from $2 million to $5 million annually.

Running the Numbers

The installed cost of residential solar photovoltaics in San Francisco is close to $10 per Watt of capacity. A typical two-kilowatt system on a single-family home costs around $20,000 before rebates and incentives. The California Solar Initiative (CSI) rebate on a system this size is $2.20 per watt or $4400 on the system as of April 2008, and the federal tax credit is $2000.1 Incorporating the $3000–$5000 San Francisco incentive, the net cost will decline to $8,600– $10,600, or about half of the original cost.

CSI rebates decline over time according to the amount of capacity put into service, as shown in the table below.2 The highlighting shows the steps within PG&E’s service territory at which rebates are being offered for the week beginning 14 April 2008. California’s three investor-owned utilities (PG&E, SCE, SDG&E) are the administrators of the CSI designated by the program’s regulator, the California Public Utilities Commission (CPUC).

One-time (‘EPBB’) rebate for small systems Performance-based incentive (PBI) for large MW in (< 100 kW, per kW installed) systems (> 100 kW, per kWh) Step step Gov’t. / Gov’t. / Residential Commercial Residential Commercial Non-profit Non-profit 1 50 n/a n/a n/a n/a n/a n/a 2 70 $2.50 $2.50 $3.25 $0.39 $0.39 $0.50 3 100 $2.00 $2.20 $2.95 $0.34 $0.34 $0.46 4 130 $1.90 $1.90 $2.65 $0.26 $0.26 $0.37 5 160 $1.55 $1.55 $2.30 $0.22 $0.22 $0.32 6 190 $1.10 $1.10 $1.85 $0.15 $0.15 $0.26 7 215 $0.65 $0.65 $1.40 $0.09 $0.09 $0.19 8 250 $0.35 $0.35 $1.10 $0.05 $0.05 $0.15 9 285 $0.25 $0.25 $0.90 $0.03 $0.03 $0.12 10 350 $0.20 $0.20 $0.70 $0.03 $0.03 $0.10

Notwithstanding the decline in rebates, the number of solar-power installations has been increasing exponentially within San Francisco since the California Energy Commission began gathering data in 1998. In that initial year, the CEC recorded three solar installations, the number growing to around 140 year in each of 2006 and 2007. By the end of 2007, 618 installations had been completed within the city, for a total of 2 megawatts. These totals exclude municipal projects.

The objective of the CSI is to stimulate California’s solar market, with a goal of one million solar roofs over ten years from the beginning of 2007. By proportion of population, this goal would equate to 22,850 solar roofs within San Francisco.3 The San Francisco Electricity Resource Plan of 2002 suggested a within-city target of 50 megawatts of solar capacity by 2012. Taking the average size of solar installations within San Francisco from the CEC data—which is around 3.2 kW, encompassing both residential and commercial systems—the 50-MW target in the ERP would equate to around 15,625 solar installations city-wide.

Using an average of the year-on-year growth rates in solar installations over the past five years (ignoring the early years because of small sample size), we can project the growth of rooftop solar

1 The federal tax credit is scheduled to expire January 1, 2009. A number of legislative efforts directed toward its renewal/extension are pending. 2 CSI Handbook, page 7. 3 California’s population is around 35 million. San Francisco’s is around 800,000. 4/14/2008 Responses_B&FCommittee_20080416.doc 3 within San Francisco under the CSI. We take the lower of the two estimates of the total potential number of solar roofs within the city (15,400) as the ‘carrying capacity’ which this growth will approach.

The GoSolarSF Incentive Payment Effect A $3 million budget for the San Francisco Solar Energy Incentive Program (SEIP) we estimate will fund 707 installations per year, net of administration costs. This estimate assumes around 4 installations are completed on commercial buildings in a year, each receiving a $10,000 incentive payment, with the balance on residential buildings, each receiving on average a $4,000 incentive payment. This is the estimated number of additional projects that could be stimulated by the San Francisco program.

The graphs below shows the possible effects of the San Francisco program. Using state data on solar installations city-wide up to the end of 2007, we project the installation rate forward in two ways, incorporating the effect of the San Francisco program in each.

The first projection assumes simply that the number of installations will climb towards a ceiling representing the availability of economically available roof space. Implicit in this projection is an assumption that either incentives are plentiful and long-lasting or that the solar industry becomes self-sustaining, by virtue of falling costs. We use a ceiling of 50 MW installed, which is the suggested target for rooftop solar by 2012 in the City’s 2002 Electricity Resource Plan. Out of this 50 MW, a portion will be attributable to the San Francisco incentive.

In the second projection, the declining availability and value of state rebates limits the rate of solar installations. The assumption here is that solar power without rebates will remain significantly more expensive than grid power so, as CSI rebates decline, so will the solar installation rate. The Appendix describes the methods in detail.

Simple projection of solar installations in San Francisco with and without SF incentive.

16,000 With SF solar incentive (max) Without SF solar incentive 14,000 ons i

t 12,000 lla a t 10,000 ins r

la 8,000 o s e

v 6,000 i t a l

u 4,000 m u C 2,000

0 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

4/14/2008 Responses_B&FCommittee_20080416.doc 4

Rebate-limited projection of solar installations in San Francisco with and without SF incentive.

9,000 With SF solar incentive (max) Without SF solar incentive 8,000

ons 7,000 i t lla

a 6,000 t

ins 5,000 r la o 4,000 s e v i t 3,000 la u

m 2,000 u C 1,000

0 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

Under the simple projection, the maximum number of additional citywide installations stimulated by the San Francisco incentive evens out at 3535 out of a total of 13,889, or about one-third. Under the rebate-limited projection, the San Francisco incentive stimulates almost six times as many additional installations (7687 compared to 1324) by the end of 2016.

Between these two scenarios lies a middle ground where state rebates will remain available at low amounts almost to 2016, as the CSI originally intended, but also where the cost of solar will decline as the market matures—if not to grid-parity then perhaps to within a fifty percent premium. In all likelihood, the San Francisco incentive will have a pronounced effect on the citywide solar installation rate, perhaps increasing it by as much as a factor of two. Building owners who might otherwise choose not to do rooftop solar because of the capital expense—even with state rebates—may find it within their grasp when the San Francisco incentive is factored in.

A further stimulus may come from Washington DC. The latest version of the energy bill passed by the Senate contains a modification to the federal tax credit for solar installations on residential buildings, which is to remove the $2000 cap. This means that residential system owners may claim a FTC of 30% of the system cost, which translates into additional savings of several thousand dollars.

Specific questions Below are responses to specific questions raised in the Budget & Finance Committee meeting of April 2nd, 2008.

Why are there so few solar installations in San Francisco relative to other Bay Area counties? Fog is not the culprit. San Francisco’s solar régime is almost as good as Sacramento’s on an average basis. San Francisco presents a few impediments to installing solar, one being the relatively high cost of doing business within the city and another being somewhat more challenging physical access to rooftops in many instances.

One of the greatest barriers to the development of rooftop solar within San Francisco is its high rental percentage. A residential tenant, although often the customer of record associated with a

4/14/2008 Responses_B&FCommittee_20080416.doc 5 meter, does not have a strong incentive to install solar power—no matter how much they may support it in principle—because they do not own the building. Landlords have few incentives to install solar power because it is tenants who pay the bulk of the electricity costs, and the common areas of rented buildings typically have little electrical load. However, the SEIP might provide tenants with leverage because an incentive is available for each and every meter within a building.

For example, a 3-kW system, which probably would be adequate for a four-unit apartment building, would cost around $30,000 before rebates, perhaps a little more if a supplemental whole-building meter were required to meet the requirements of the CSI, or if power from the system were routed to more than one inverter. Conservatively, assume the gross cost to be the total cost for a 3 kW photovoltaic system is $34,000. CSI rebates would be $6600 and the federal tax credit would be $2000, assuming a single inverter for the system. Without the San Francisco incentive, the net cost, then, would be $25,400, which the landlord would have to pay. Their only way to recoup this cost from sitting tenants would be to increase their rent, which would be possible only with a tenant’s agreement. Tenants might agree to this knowing their PG&E bill would diminish by a corresponding amount, or if they planned to remain in the same place for a long period; however, the more apartments in a building, the less the likelihood of unanimity among the tenants.

The San Francisco incentive would pay a minimum of $3000 per meter, which equates to $15,000 for the five meters in a four-unit apartment building. Now the net cost of the system is $10,400. At this price, some of the more forward-thinking landlords might be motivated to do solar purely on the basis of its adding to the value of a property. The most likely scenarios in which landlords may take advantage of the SEIP are while renovating a property for rent or for properties that have high turnover.

How have MECA funds been used? What are some of the alternative uses of the funds, and are other uses a better investment? Each year, through the budget process, any HHWP net operating revenues are appropriated toward various uses. Some examples include power house improvements and specific streetlighting improvements. In 2001, in response to California’s energy crisis, Mayor Brown established the Mayor's Energy Conservation Account to fund renewable generation and energy efficiency projects. To-date, $76,914,926 has been appropriated to MECA from Hetch Hetchy net operating revenues.

Energy efficiency allocations are $30,672,955, with renewable-power allocations of $38,918,198.4 Other MECA monies, totaling $7,323,773, have been allocated to CCA and Treasure Island, with the balance as yet unallocated.5

In FY07-08, $9.9 million was allocated to MECA, with $5.2 million targeted toward renewable projects and $4.7 million targeted toward energy efficiency projects. From the $5.2 million for renewable projects, $3 million was initially allocated to the 250 kW expansion of the solar installation at Pier 96. This project was to be completed under a design-build project structure, which would require the full cost of the project to be expensed at an estimated $3 million in FY07- 08. Instead, the project will be completed under a power purchase agreement (PPA) structure, together with the 5 MW Sunset Reservoir solar project. Under this structure, a developer will be selected through a competitive process. The developer will design, build, own, operate and maintain the solar systems, leasing the roof space from the City, and sell all electricity produced from the solar systems to the PUC, which shifts any performance risk to the developer. The developer will be able to access the federal tax credit, amortize the depreciation, and in so doing lower the cost of the systems. The PUC has the right to purchase the systems after the 7th year of

4 The renewable-power project figure includes a $12 million spending authorization for FY 03–04 where COPs were expected to be issued. They were not. 5 The Appendix 1 provides the specific breakdown of costs for each project. 4/14/2008 Responses_B&FCommittee_20080416.doc 6 operation or at any point thereafter at some cost. The PUC’s annual cash outlay for the output of the systems is estimated at $2 million. The Sunset Reservoir and Pier 96 solar power purchase agreements are currently out for bid, with bids due on April 11, 2008. The PPA structure, being less capital-intense than the design-build approach, will allow the SFPUC to allocate the excess funds to the SEIP.

The $3 million SEIP provides an opportunity to leverage much more installed capacity among residents and businesses than a design-build municipal project. Assuming the program funds the installation of 720 two-kilowatt systems per year, the total output of those systems would be just over 2000 megawatt-hours per year.6 A municipal project costing $3 million would have an installed capacity of around 250 kW.7 Based on data from the installation at the Moscone Conference Center, a system of this capacity would produce around 270 MWh per year.

The municipal project produces revenue, however. Municipal customers are charged the same as their normal Hetch Hetchy rates, either zero, $0.0375, or about $0.12 per kWh. Assume the rate is $0.12 per kWh, then annual revenue offsetting the $3 million capital cost will be $32,400.8

Residential and commercial systems funded in part by the SEIP will not generate direct revenue to the City. Revenue will accrue indirectly over time, as properties change hands and their value is reassessed. Rooftop solar is expected to increase property values, but by how much is not yet well known because the sample of relevant transactions is still small.

Should the funds be used for solar panels or for energy conservation? Should they be used for public properties or for public outreach and support for solar installers? What is the megawatts-to-dollar ratio in solar, energy conservation and refrigerator recycling? The CSI contains a built-in requirement for a customer to undergo an energy efficiency audit as one of the conditions for receiving a solar rebate. For most residential customers, this audit is comprised of participating in an online survey. Commercial customers are subject to more stringent requirements.

The SEIP is structured to follow the CSI, which means that in order to receive a San Francisco incentive a customer must pass through the energy efficiency audit. Furthermore, PG&E’s interim marketing budget of $500,000 annually for the CSI is pending CPUC review. If approved, CSI program administrators would be allowed to spend up to 5 percent of their total CSI funds on administration, which includes marketing and outreach. In PG&E’s case, this amount would equate to around $42 million.9 While not wanting to duplicate PG&E’s effort, the SEIP intends to inform customers of the range of energy efficiency programs available to them through PG&E, as well as the programs available through the San Francisco Department of Environment (SFE).

SFE administers two energy efficiency programs. The Small Business Program offers small businesses a direct-install service package, mostly for lighting and HVAC, which includes audit, design and construction. SFE reports that the program has $5 million in funding for three years, with goals of 20 million kWh saved at an average cost of around $0.20 per kWh.

The second SFE program is the Plus Program, which is for commercial multi-family buildings, covering common areas, such as hallways, laundry and exterior lighting. SFE reports that this program has $6 million in funding for three years, covering lighting, controls and energy-efficiency

6 720 × 2 kW × 1400 kWh per kW per year. 7 Municipal projects are not eligible for state rebates because they are not retail electric customers of PG&E and therefore do not pay the Public Goods Charge that funds the CSI. 8 270,000 kWh × $0.12 / kWh. 9 Cmr. Peevey’s Ruling requesting comments on the interim marketing and outreach plans for the CSI, dated 4/3/07, follows from the original administrative budget set in CPUC Decision 06-12-033, December 14, 2006 (Rulemaking 06-03-004). 4/14/2008 Responses_B&FCommittee_20080416.doc 7 processes, with a goal of 35 million kWh saved at an average cost of $0.10 per kWh. The program is vendor-driven: vendors are given incentives for marketing projects.

Applying the SEIP funds to a new energy efficiency program could also achieve electricity savings. PG&E administers a number of appliance rebate programs which offer from $30 to $75 dollars when installing a high efficiency appliance. For most residential customers in San Francisco, the highest energy consuming appliance is the refrigerator. PG&E offers a $35 rebate to haul away an old fridge.10 Replacing an older fridge with a new, Energy-Star-certified one will save an average of 400–700 kWh per year, at a refrigerator purchase price of $400–$700. If the $3 million SEIP budget were used to offer San Francisco residents and businesses a replacement fridge in this price range (at an average of around $500, net of the PG&E rebate) then 6000 fridges could be replaced, which would save a combined 2,400–4,200 MWh per year, or somewhat more than the 2,000 MWh annual savings in grid power from 707 rooftop solar installations. Alternately, if the SEIP budget were used to match PG&E’s $35 rebate then a maximum of 85,700 fridges could be replaced for a combined savings of 34,280-59,990 MWh per year. This total assumes that 85,700 fridge owners within San Francisco then would be committed to spending several hundred dollars on a new fridge, which is probably unrealistic.

While the energy savings from replacing old fridges are better than from solar panels, stimulating the local solar market produces other benefits which fridge-recycling does not, most notable of which is an increased demand for local, skilled installers. Beyond the benefits already being felt in Bayview-Hunters Point by the training of new installers made evident in public testimony, a proliferation of small rooftop solar projects around the city will keep a small army of installers busy, potentially for years to come, if the program is made permanent. These workers will generate far more in payroll taxes to the City than the contractors on the few, large municipal solar projects completed to-date. A large number of small installations results in a much higher aggregate labor cost per installed kilowatt than a small number of large systems. Further, the municipal solar program will benefit from the increased competition to provide solar installations in San Francisco.

What about the utility undergrounding program? Building owners are responsible for undergrounding utility connections from the street or easement to their buildings as PG&E, and other utility companies, underground their distribution systems. Some building owners cannot afford the cost to underground the connection to their property, which can sometimes hold up the undergrounding of an entire block. While the SEIP funds could be reappropriated to provide assistance to these homeowners, the re-purposing of a portion of the seismic loan funds proposed by Mayor Newsom may be a more appropriate source of assistance.

How many people are lined up at PG&E to do this program? Can we get a breakdown of how the state PG&E rebate program is set up? PG&E reports twenty projects within San Francisco currently in the queue to receive CSI rebates. Solar installers report that prospective customers have postponed solar projects, pending the launch of the SEIP, so that they could take advantage of the San Francisco incentive.

Beyond the CSI rebate amounts displayed in the table above, the CSI process is a two-step procedure for most applicants. In the first step, in advance of installation, an incentive may be reserved from the CSI budget provided that an application meets various validity and system design criteria. In the second step, an incentive is claimed after a system is interconnected to the PG&E grid. A customer must submit an incentive claim form, which is reviewed by PG&E, who then issues a CSI Approval of Incentive Payment. A check follows in the mail within 30 days.

10 http://www.pge.com/myhome/saveenergymoney/rebates/recycling/index.shtml 4/14/2008 Responses_B&FCommittee_20080416.doc 8

To minimize paperwork, the SEIP is structured to tie into the CSI. A customer can apply for and reserve a San Francisco incentive by showing proof of a CSI incentive reservation. By providing the CSI Approval of Incentive Payment, a customer later can claim a San Francisco incentive payment.

What are the other public sites where we could develop projects under the arrangements where we leverage funds (like at Sunset Reservoir & Pier 96) but keep a closer, public control over the assets? With the passage of the California Solar Initiative law in 2006, and the implementing decisions by the California Public Utilities Commission, the City can no longer receive a rebate when it develops a solar generation facility. That’s because the CSI requires recipients to be a retail electric customer of an investor owned utility. These changes in the CSI program encouraged PUC to explore other approaches to developing solar generation facilities. In particular, PUC explored the power purchase agreement structure to developing solar generation facilities. Under this model, the developer of the solar generation facility, while still not eligible for the CSI rebate, is eligible for the federal tax credit and accelerated depreciation. The PPA contains a variety of options for allocation of ownership. SFPUC can take ownership of the projects as early as year seven. The PPA structure reduces project costs because the developer, being a private entity, is eligible to take the federal tax credit (30% of system cost, which is significant) as well as accelerated depreciation, instruments not available to public entities.

Below is the list of potential projects identified by the PUC for development through the PPA structure:

Phase 1 2008 (6MW) Sunset Reservoir – North Basin 6MWp Pier 96 0.3MWp

Phase II 2008/09 (36MW) Stanford Heights Reservoir 1.3 MWp SFGH Parking Garage, 24th & Utah 0.5 MWp Cable Car Barn 0.2 MWp Bus Washing Facility 15th & Harrison 1 MWp Tesla, Ground-mounted 7 MWp Sunol, Ground-mounted 26 MWp

Phase III 2010 - 2012 (16MW) University Mound - North Basin (2010) 2 MWp Pulgas Reservoir (2010) 2.6 MWp Sutro Reservoir (2011) 2 MWp Hunters Point (Parcel E) Ground-mounted 10MWp

All of these sites are city-owned, and any solar system developed on the site under the PPA approach would be owned by the PUC as early as year seven of the contract, at some cost to the PUC, or as late as year twenty-five, at no additional cost. Whereas design-build projects cost the PUC around $10 per Watt installed, projects funded via PPA generally cost the equivalent of around $5–$7 per Watt. However, the actual cost will depend upon a variety of factors, such as project size, availability of tax credits, and the financial pro forma employed by the development company.

Why can’t the PUC use bonds to fund these sorts of solar projects, like the voters approved? The PUC’s Power Enterprise is working with outside counsel to position itself to access the capital markets. In order to accomplish its mission the Power Enterprise needs to be a viable and sustainable business with a financial strategy that generates the resources sufficient to fund:

4/14/2008 Responses_B&FCommittee_20080416.doc 9

• Ongoing operations and maintenance expenses • Replacement and reconstruction of existing capital assets • Additions to or extensions of its capital asset base.

A condition precedent to accessing cheap capital is a credit rating. The Credit Analysis Framework table below sets forth the basic framework that credit rating agencies use to evaluate the credit- worthiness of public power agencies. The key areas are management and business strategy, rates, regulation, generation and distribution facilities, customer demographics, financial performance, risk management practices, and legal provisions relating to public offerings of debt.

Credit Analysis Framework

Management and Business Strategy Demographics • Description of structure and business units. • Population and customer growth rates; • Management and board structure. usage trends. • Working relationship between management, the • Income levels, employment trends, and board, and members, among others. unemployment level. • An assessment of the competitive strategies • Customer relations. (short- and long-term). • Service area franchise provisions. • Economic development. Rates Financial Performance • Rate design and pricing strategies. • Five-year historical financial summary. • Rates by customer class and actual rates for • Financial projections. key customers. • Capital requirements. • Usage of longer-term customer contracts. • Financial policies and management targets. • Future projections. • Policy on short-term debt usage. • Rate comparison with other regional suppliers. • Debt retirement or restructuring strategies. • Wholesale systems — review of rates for power supplier and distribution systems.

Regulation Risk Management Practices • Ratemaking body. • Fuel supply • Rate-setting procedures and history. • Wholesale market activity • Level of involvement by federal, state, or local • Hedging policies agencies in utility affairs. • Financial liquidity resources

Generation and Distribution Facilities Legal Provisions • Review of generating facilities and how they • Bond covenants. meet system load requirements. • Debt service reserve policy. • Discussion of the regional power market and the • Outstanding litigation. utility’s position. • Contractual obligations. • Potential alliances. • New projects under consideration. • Fuel contracts and transportation agreements. • Fuel price volatility hedging practices. • Transmission and distribution issues. • Plant production costs. • Environmental issues.

How do we know the pilot program is really adding solar systems that would not have otherwise been built? What are our criteria for measuring success? Even in pilot phase, we have ways to measure results, as the bar chart above illustrates. Above we have provided a credible baseline from which to measure performance. With that baseline, it

4/14/2008 Responses_B&FCommittee_20080416.doc 10 will be possible to measure how many additional projects are stimulated by the San Francisco incentive.

Means testing, versus a free-for-all? Installing a solar generation facility, even with the proposed SEIP, state rebate and federal tax credit available to private property owners, requires access to thousands of dollars. Under the typical system example given above, a property owner would still have to come up with between $8,600-$10,600 (or $11,600-$15,600 without the proposed SEIP). Many of the public comments in the meeting stressed the benefits of beginning the program as soon as possible in pilot form, so as to take advantage of the highest-available state rebates, even if not all of the desired provisions might be in place before the program is made permanent.

Two other pending program efforts, together with the proposed SEIP, could make investment in solar reachable by a greater percentage of San Franciscans: low-cost loans, proposed by Mayor Newsom, and the assessment concept, proposed by Supervisor Sandoval. The prospect of low- cost loans for solar and energy efficiency improvements could make the investment more attainable to “house-rich, cash-poor” property owners. This proposal would re-purpose untapped bond capacity approved for a seismic retrofit loan program. Likewise, it could also be helpful to have the option to repay such a loan by opting into an assessment district. Loan repayment would appear on the tax bill of the property owner. Under this proposed program, obligation to repay the loan for the solar system would transfer to a new property owner at the time of sale of the property.

The state rebate program does not currently have a low income component, though the California Public Utilities Commission is developing such a program. The proposed SEIP could incorporate a means test. PUC has experience implementing means-tested programs in its role as San Francisco’s water and wastewater utility. The Community Assistance Program (CAP) for water and wastewater customers provides a discount to residential, single-family, low-income customers who qualify and complete a one-page application form. PUC could build upon this information to develop a low income component to the SEIP program.

What are the cost containment measures? A concern exists whether solar installers might simply increase their prices in line with the incentive amounts. Competition among Bay Area solar installers is intense. The solar installation market acts a regional market and is compromised of solar installers from San Jose to Sonoma county. Competition among installers is expected to ensure the benefit of the incentive payment accrues to the property owner in an overall lower installation cost.

During the public comment period, Nancy Pfund mentioned “Brightsource’s deal with PG&E” and “Powerlight in Richmond”. Please provide details. Ms. Pfund was referring to a supply contract Brightsource Energy has signed with PG&E to supply 500 MW from three large solar thermal plants in the Mojave Desert, with an option to purchase an additional 400 MW. The agreement will help PG&E meet its state-mandated requirement of obtaining 20 percent of its energy from eligible renewable sources by 2010. See www.brightsourceenergy.com along with an April 1 article in the SF Chronicle. Brightsource Energy is headquartered in Oakland.

Ms. Pfund’s reference to “PowerLight in Richmond” is PowerLight’s absorption by SunPower, coincident with SunPower’s establishment of a new headquarters in Richmond at the old Ford Motor Plant. See www.sunpowercorp.com.

4/14/2008 Responses_B&FCommittee_20080416.doc 11

Appendix 1 MECA projects expenditure summary.

Status Available FY Created Number Project Title Budget Spent (Y/N/Cancelled) Balance 01-02 CUH947 01 N Master MECA 4,731,191 1,148,246 409,445 01-02 CUH947 02 Y Energy Efficiency Fair 93,133 93,133 - 01-02 CUH947 03 Y SFGH Lighting System Retrofit 11,472 11,471 - 01-02 CUH947 04 Y DPH Clinics 1,325,630 1,325,630 - 01-02 CUH947 05 Y LED Traffic Signals (6,758) (6,758) - 01-02 CUH947 06 Y Refrigerators - Housing Authority 1,062,038 1,062,038 - 01-02 CUH947 07 Y SFGH Energy Efficiency & Generation 150,918 150,973 (55) 01-02 CUH947 08 Y REC & Park Energy Efficiency 76,415 76,415 - 01-02 CUH947 10 Y Moscone Renewable Energy/Energy Efficiency 7,936,740 7,936,740 - 02-03 CUH947 11 Y Lighting Retrofit BBR 45,999 45,999 - 02-03 CUH947 12 Y DPT Garages 138,583 114,963 23,620 02-03 CUH947 13 C Fire Stations 467 467 - 02-03 CUH947 14 Y West Portal Library Lighting 20,335 20,335 - 02-03 CUH947 15 C 1155 Market LEED Certification 94,707 94,707 - 02-03 CUH947 16 C Demand Management Program 295,214 228,429 66,785 02-03 CUH947 18 Y Moscone West Energy Efficiency 5,500,000 5,500,000 - 02-03 CUH947 19 Y WPC - SEP Energy Efficiency 34,383 34,382 - 02-03 CUH952 01 Y MECA - SFGH HVAC 1,350,000 1,290,923 59,077 02-03 CUH954 01 Y MECA - SFPUC High Energy Users (Moscone West) 1,500,000 1,500,000 - 02-03 CUH955 01 Y MECA - CWP Southeast Wastewater Solar 2,075,000 1,933,767 141,233 02-03 CUH955 02 Y Solar Monitoring 25,000 23,445 1,555 03-04 CUH947 20 N Moccasin Energy Efficiency 802,645 161,319 641,326 03-04 CUH947 21 N Load Monitoring and Controls 468,175 420,949 47,226 03-04 CUH947 22 C Port of SF Energy Efficiency 2,041,000 508,954 1,532,046 03-04 CUH947 26 Y Moscone West Recommissioning 63,791 62,930 861 03-04 CUH960 01 Y Solar Power Program 12,842,782 294,286 48,496 03-04 CUH961 01 Y MECA - Solar Project 795,000 689,624 105,376 04-05 CUH947 27 N WPC/CDD Energy Efficiency/Demand Management 500,000 220,950 279,050 04-05 CUH947 28 N MUNI Energy Efficiency 496,931 441,863 55,068 04-05 CUH947 29 N SF Airport Energy Efficiency 1,758,422 1,515,639 242,783 04-05 CUH960 04 Y Solar at Pier 96 2,700,183 2,609,539 90,644 04-05 CUH965 01 N Solar Project - Airport / MUNI / Port - 110 (110) 04-05 CUH966 01 Y MECA - Demand Reduction SFPUC 705,000 674,058 30,942 05-06 CUH947 31 Y Tidal Flow Power Feasibility Study 96,140 96,140 - 05-06 CUH947 32 C Solar Project - SF General Hospital 102,025 102,025 - 05-06 CUH947 33 Y Solar Project - NorthPoint 2,992,722 2,619,511 373,211 05-06 CUH947 34 Y Solar Project - CDD 1,534,354 1,379,296 155,058 05-06 CUH947 35 C Solar Project - Pier 50 60,366 60,366 - 05-06 CUH947 36 Y Solar Project - SFO 6,256,000 6,208,469 47,531 05-06 CUH947 37 Y 10 Site Solar 845,820 918,638 (72,818) 05-06 CUH947 38 N Solar Project - SF MUNI 1,044,058 38,330 1,005,728 05-06 CUH947 39 Y Solar Project - SF Libraries 40,343 40,343 - 05-06 CUH947 40 N SFGH Central Plant / HVAC 521,169 12,865 508,304 05-06 CUH947 41 N SFGH Chiller Replacement 1,800,000 1,800,000 - 05-06 CUH947 42 N Treasure Island Redevelopment 108,153 18,116 90,037 05-06 CUH947 44 N Energy Efficiency - General Fund Depts 356,000 117,064 238,936 05-06 CUH947 45 N New Construction Design Review 438,577 275,236 163,341 06-07 CUH947 46 C Fuel Cell - HHP 1,742 1,742 - 06-07 CUH947 47 N Asian Art Museum Energy Efficiency 6,433 6,433 - 06-07 CUH947 48 N Wastewater Energy Efficiency 334,000 148,612 185,388 06-07 CUH947 49 N Treasure Island Submarine Cable 2,599,340 2,599,340 - 06-07 CUH947 50 N Renewable Project Planning 500,000 230,194 269,806 06-07 CUH947 51 N MUNI Woods Coach 1,636,000 3,735 1,632,265 06-07 CUH947 53 N Geothermal Project 280,000 63,361 216,639 06-07 PUH501 01 N SFE Energy / Green Building 465,000 - 465,000 06-07 PUH501 02 Y Tidal Feasibility Studies 85,123 85,123 - 07-08 CUH947 54 N Compact Fluorescent Lamp Program 171,625 79,648 91,977 07-08 CUH947 55 N New Site Study (Library/Hunter's Point) 200,000 8,251 191,749 07-08 CUH947 56 C Pier 96 Expansion 10,000 13,156 (3,156) 07-08 CUH947 57 N Design Build Contract Development 505,540 19,348 486,192 07-08 CUH947 60 N Sava Pool Solar System 200,000 - 200,000 07-08 CUH947 61 N CEC Geothermal 500,000 - 500,000 07-08 CUH947 62 N In Line Turbine 290,000 - 290,000 07-08 CUH947 63 N Solar Energy Incentive Fund 3,000,000 - 3,000,000 07-08 CUH947 64 N Chinatown Health Center 300,000 - 300,000 TOTAL: 7 6,914,926 47 ,130,868 14,11 0,556

4/14/2008 Responses_B&FCommittee_20080416.doc 12

Appendix 2 How solar installations are projected.

Data sources, assumptions & method Annual solar installations within San Francisco to the end of 2007 are estimated by combining data from three California state datasets: • CEC dataset of projects funded under the Emerging Renewables Program (ERP) and New Solar Homes Partnership (NSHP)11; • CPUC dataset of projects funded under the Self-Generation Incentive Program (SGIP), excluding municipal projects12; • CPUC dataset of projects funded under the California Solar Initiative (CSI).13

These datasets combine to produce annual and cumulative city-wide installation trends shown in the bar charts below.

180 168

160 152

s 140 n o i

t 117 a

l 120 al

st 100 n

i 85 r a l 80 so

al 60 u 45 n

n 38

A 40 21 20 3 66 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

700 641

600 s n o i t 500 473 alla st 400 in

lar 321 o

s 300 236 ve i

lat 200 u

m 119 u

C 100 74 36 3 9 15 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

11 http://www.energy.ca.gov/renewables/emerging_renewables/index.html 12 http://www.pge.com/b2b/newgenerator/incentive/availablefundingandprogramstatistics.shtml 13 https://csi.powerclerk.com/Default.aspx 4/14/2008 Responses_B&FCommittee_20080416.doc 13

In the ten years from 1998 to 2007, a total of 641 projects have been completed within San Francisco for a total installed capacity of just over 2.6 megawatts.14 These projects have benefited from a total of $8.8 million in rebates from the various programs, for an average of around $3600 per kilowatt or slightly under $13,000 per project. As the table below shows, residential and commercial projects differ in size and in the amount of rebates they receive.

Non-municipal rooftop solar Residential & Small commercial Large commercial All projects in SF, 1998–2007 ERP, NSHP, CSI SGIP Total projects, 1998–2007 633 8 641 Average installed capacity (kW) 3.2 71.7 3.6 Average rebate per project $10,154 $295,869 $12,902 Average rebate per kW $3,125 $4,125 $3,584

There is no single ‘correct’ way to project rates of solar installations into the future. Installation rates will depend on incentive amounts, costs of solar technology and general economic conditions, to name a few. In any case, we assume the number of solar installations will increase towards a theoretical plateau, which reflects constraints in the availability of incentives or, in the limit, of economically available roof space. This assumption results in an S-shaped growth curve which can be produced by the following discrete growth function,

⎛ K − Nt ⎞ Nt +1 = Nt + rNt ⎜ ⎟ ⎝ K ⎠ where Nt+1 is the cumulative number of solar installations in year t+1; Nt is the cumulative number of solar installations in year t; r is the year-on-year growth rate in solar installations; K is the theoretical plateau of installations within the city (also called a carrying capacity).

Early growth is not limited by K, so it is exponential but, as N approaches K, r tends to zero.

The historical trend illustrated in the bar charts above points to the direction of future growth, allowing us to estimate a future growth rate. The last five years of growth are used to estimate an average year-on-year growth rate, thereby avoiding any sample bias from the early years, which contain small numbers of installations. The number of installations by the end of 2007 is a function of the number of installations at the end of 2002, the year-on-year growth rate, r, and the number of years, t = 5.

t N 2007 = N 2002 ()1+ r which resolves to

1 ⎛ N 2007 ⎞ t r = ⎜ ⎟ −1. ⎝ N 2002 ⎠

The average year-on-year growth rate this function produces is 54%.

We illustrate the possible effect of the San Francisco solar incentive by first projecting annual installations without the incentive, and then we incorporate a maximum number of annual

14 A handful of projects were in place before 1998, the first year of the datasets listed above. 4/14/2008 Responses_B&FCommittee_20080416.doc 14 installations funded by the incentive. This maximum number assumes a constant incentive budget of $3 million, renewed annually. Of this budget, $150,000 is allocated annually to program administration, leaving $2.85 million per year for incentives.

The maximum number of additional installations stimulated by the San Francisco incentive is estimated by assuming four commercial installations per year at $10,000 each, with the remainder being residential at an average incentive of $4000 per system. These assumptions produce a total of 707 installations per year. Note that this is a maximum: the actual number of additional installations stimulated by the San Francisco incentive will lie somewhere between the baseline amount and this maximum.

Of the two projections that follow, the first is a somewhat optimistic baseline projection of installations and the second a somewhat pessimistic one. We expect the actual number of future solar installations within San Francisco to lie somewhere between the two.

Simple projection In the simple projection, rebates are assumed plentiful and long-lasting, so the growth in installations is limited by economically available roof space. The size of this limit is taken from the 2002 Electricity Resource Plan, which suggests a target of 50 megawatts of rooftop solar capacity by the end of 2012. At an average size of 3.6 kilowatts per system, this 50 MW would equate to 13,889 solar installations.

The 641 total solar installations at the end of 2007 is projected forward using the growth function above, with r set to 0.54 and K set to 13,889. The 707 SEIP-induced installations are included in the total. As the cumulative number of installations approaches 13,889, the installation rate slows to zero. The projection encompasses all future years of the California Solar Initiative, out to 2016. From the cumulative projection, we infer the annual totals by subtraction.

4,500 With SF solar incentive (max) Without SF solar incentive 4,000

3,500 ons i t 3,000 lla a t 2,500 ins r

la 2,000 o l s 1,500 nua n 1,000 A

500

0 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

4/14/2008 Responses_B&FCommittee_20080416.doc 15

16,000 With SF solar incentive (max) Without SF solar incentive 14,000 ons i

t 12,000 lla a t 10,000 ins r

la 8,000 o s e

v 6,000 i t a l

u 4,000 m u C 2,000

0 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

Under this projection, the ceiling of 13,889 citywide installations is quickly achieved, by coincidence, in 2012. In reality, of course, the annual installation rate will never actually be zero but the purpose of the model is to illustrate the possible effect of the San Francisco incentive. In the simple projection, the San Francisco incentive has a pronounced but not overwhelming effect on total installations.

Rebate-limited projection In the rebate-limited projection, we constrain growth in installations to the period during which state rebates remain available. By far the largest state program is the CSI, so we use the trend in the CSI budget as a proxy for the availability of all incentives.

The CSI dataset includes rebate amounts paid and reserved for every individual project. Reservation and payment dates also are included. Using these data, we can estimate how fast the CSI budget within PG&E’s service area is being used up. Unlike previous state programs, whose budgets were renewed every year, the CSI has a ten-year budget, to encourage the solar industry to plan for the long term. As the graph below suggests, however, the budget is going fast. The demand for rebates has been so strong that, out of ten stepwise reductions in rebate amounts planned over the ten years of the CSI, rebates for residential systems already are at step 4 and those for non-residential systems are at step 5.15 And we are only 16 months into the program.

15 See the Statewide Trigger Tracker at http://www.sgip-ca.com/ 4/14/2008 Responses_B&FCommittee_20080416.doc 16

$1,000

$900

$800

t $700

$600 budge I

S $500 C

E $400 & G

P $300

$200

$100

$- Dec-06 Mar-07 Jun-07 Sep-07 Dec-07 Mar-08

PG&E’s budget is being used up at over $200 million per year, and its budget at the end of 2007 was just over $600 million, so at that rate the budget will be all gone by the end of 2010.

The rebate-limited projection adjusts the annual growth rate in solar installations in proportion to the decline in the rebate budget. Installation rates will decline as the budget is used up because, as more capacity is installed statewide, the lower rebates will become, which reduces a building owner’s incentive to do solar. However, the decrease in installation rates will also cause the budget itself to decline at a slower rate. The linear decline we see in the graph above then will become a more of a logarithmic decline as in the bar chart below.

800

700 on ti

c 600 e ) oj ea

ar 500 t pr e c i v dge r 400 e s bu I E S

& 300 C G r P a ( e

y 200 - d i

M 100

0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

The projection in the bar chart incorporates the $200 million-per-year decline in PG&E’s CSI budget up to 2009 and then incorporates a more gradual approach to zero by dividing the previous year’s number by two. This budget projection is used to adjust the San Francisco annual growth rate in installations, by proportion.

The rebate-limited projection assumes an absence of major technological improvements in solar technology that could lead to the cost of solar approaching grid-parity. The intent of the CSI is to stimulate a self-sustaining solar market in California, which would imply a per-kWh cost not very

4/14/2008 Responses_B&FCommittee_20080416.doc 17 far above grid power. At present, it is more than twice as expensive as grid power on a lifetime basis. Acknowledging that new, disruptive technologies could radically improve the economics of solar power during the lifetime of the CSI, the rebate-limited projection takes a conservative approach, assuming that solar remains expensive.

On these assumptions then, as the rebates decline towards zero, the growth rate in installations, r, also will decline towards zero. As in the simple projection, we calculate cumulative installations by year and then infer annual totals by subtraction. The San Francisco incentive effect is then added on top.

1,200 With SF solar incentive (max) Without SF solar incentive 1,000 ons i t 800 lla a t ins

r 600 la o l s 400 nua n A 200

0 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

9,000 With SF solar incentive (max) Without SF solar incentive 8,000

ons 7,000 i t lla

a 6,000 t

ins 5,000 r la o 4,000 s e v i t 3,000 a l u

m 2,000 u C 1,000

0 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

In the rebate-limited projection, the cumulative number of installations without the San Francisco incentive reaches a plateau of around 1300 city-wide by the end of 2013, as contrasted with more than 10,000 under the simple projection.

After 2013, with state rebates having been used up, the model assumes the only installations being done locally will be those under the San Francisco incentive.

4/14/2008 Responses_B&FCommittee_20080416.doc 18

Discussion Under the simple projection, the maximum number of additional citywide installations stimulated by the San Francisco incentive evens out at 3535 out of a total of 13,889, or about one-third. Under the rebate-limited projection, the San Francisco incentive stimulates almost six times as many additional installations (7687 compared to 1324) by the end of 2016.

4/14/2008 Responses_B&FCommittee_20080416.doc