Comments on 2018 Virginia Energy Plan

Dr. Damian Pitt, Associate Professor and Program Chair

Urban and Regional Studies and Planning

L. Douglas Wilder School of Government and Public Affairs

Virginia Commonwealth University

My comments below are based on the expertise I have gained through 12 years of research on the intersection of state and federal energy policy with local-level urban planning processes, particularly as it relates to distributed renewable energy. The observations and recommendations contained herein are my own, and do not necessarily represent the positions of the L. Douglas Wilder School of Government and Public Affairs or Virginia Commonwealth University.

I recommend that the 2018 Virginia Energy Plan recognize the viability and importance of distributed solar PV to the electrical grid, as well as the rights of Virginia’s residents and businesses to utilize distributed solar PV via ownership models that are common practice in many other states.

It is important to recognize that, while utility-scale solar energy is booming in Virginia, we are trailing far behind other comparable states when it comes to the deployment of distributed solar energy. Virginia had zero utility-scale facilities at the beginning of 2016, but is now the 14th leading state, with almost 370 MW, as of the end of 2017. While distributed PV has also been increasing, from just over 5 MW in 2011 to 46.5 MW by the end of 2017, Virginia still ranks only 31st in total distributed PV and 38th in distributed PV per capita. Only 11% of the state’s total PV capacity is from distributed systems, compared to 39% nation-wide, which puts Virginia 45th among all states in that regard. Our distributed PV total is far behind those of neighboring Maryland (619 MW) and North Carolina (126 MW), and barely ahead of the District of Columbia (38 MW) (U.S. Energy Information Administration 2017b, 2018b).

The benefits of distributed solar energy, particularly its contributions to pollution reduction, greenhouse gas mitigation, and economic development, are well-documented and will surely be elucidated in many other comments on the Virginia Energy Plan. Therefore, I am focusing my observations on refuting two of the criticisms of distributed solar energy that are occasionally levied by for-profit electric utilities and fossil-fuel interests: that it creates net-costs that utilities must pass on to all retail electric consumers (i.e., “cross-subsidization”) and that it creates technical problems for the electrical distribution grid.

While both of those critiques have some conceptual merit, and must be addressed in the long term, an abundance of research from across the country demonstrates that it is highly premature to use those arguments to slow down the growth of distributed solar energy today.

I have reviewed dozens of studies on these topics, published by U.S. national laboratories, academic researchers, and other qualified professionals. I have also conducted my own research, in partnership with electrical engineers at my university, using data from Virginia. Based on this research, it is clear that distributed PV capacity can be significantly raised beyond currently levels before the aforementioned costs and technical impacts become problematic.

No Evidence of Cross-Subsidization of Distributed Solar PV in Virginia

Key findings regarding the so-called cross-subsidization of distributed PV are as follows:

• Any cost impacts from distributed PV would only occur if the net benefits of the electricity produced by those systems adds up to less than the retail electric rate, AND if the market penetration of DPV systems is great enough for those costs to be significant.
• While no comprehensive “value of solar” (VOS) study has been undertaken in Virginia, dozens have been completed for other states and regions across the country. The general consensus among independent VOS studies (those completed on behalf of a Public Utility Commission rather than an electric utility or solar energy advocates) is that the net benefits provided by distributed solar are roughly equal to, if not above, the retail electric rate (see Hallock & Sargent 2015, and Barbose 2017). In that case there can be no cross-subsidization, as there are no net costs to pass on to other consumers.
• A report by the Lawrence Berkeley National Laboratory shows that even if the net VOS is believed to be lower than the retail rate, the actual cost impact (i.e., the extent of cross-subsidization) is likely negligible. This is because, given the very small market penetration of distributed PV systems in most regions, the magnitude of the costs incurred would be tiny when spread across the utility’s entire customer base.

For example, the national laboratory’s methodology shows that if the energy from distributed PV systems is worth only 75% of the retail electricity rate, which is on the low end of the typical range for most VOS studies, then a distributed PV market penetration of 5% would result in a retail rate increase of 1.25% (Barbose 2017). In Virginia, where the retail rate averages around 12 cents per kWh, this would mean a rate increase of less than two tenths of a cent per kWh.

• A 2011 study by the Virginia State Corporation Commission (SCC) shows that, even with very conservative assumptions about the value of solar, increasing Virginia’s distributed PV market penetration to 1% would only raise the average residential customer’s bill by a little over 50 cents per month.
• Applying the Lawrence Berkeley National Laboratory methodology to the Virginia SCC data suggests that, even with the conservative assumptions from the 2011 SCC study, raising the market penetration all the way to 5% would still only increase electricity rates by less than $3 per month on average. With less conservative assumptions about the value of solar, these cost impacts would be even lower.

In summary, the existing research on the value of solar clearly demonstrates that distributed PV use in Virginia can be increased to far above existing levels, without creating any significant costs to utilities or resulting cross-subsidization of distributed PV use by non-PV customers.

The Electrical Grid in Virginia can “Host” Much Higher Levels of Distributed PV

Numerous studies from various federal government agencies and research laboratories show that existing rooftops can support vast amounts of distributed PV across the country, much of which can be supported with existing electricity market conditions and grid infrastructure.

• The US Department of Energy’s SunShot Vision Study (2012) estimated Virginia’s combined utility-scale and DPV market potential to be 8,700 MW by 2030, and 21,200 MW by 2050.
• Two studies by the National Renewable Energy Laboratory (NREL) found that Virginia’s buildings could support from 21,800 MW (Paidipati et al., 2008) to 28,500 MW (Gagnon, et al., 2016) of distributed PV, the latter of which would be enough to provide nearly a third of the state’s annual electricity consumption.
• Two other NREL studies estimated that Virginia’s electrical grid infrastructure could support 19 GW of distributed PV, more than 400 times the state’s current total (Denholm & Margolis 2008, Lopez, et al. 2012).
• A study for Dominion Energy by Navigant Consulting (2016) found that eleven “representative” distribution feeders – out of a total of 1800 across the Dominion service territory – could host over 200 MW of distributed PV just by themselves. This is more than four times Virginia’s current state-wide capacity. 

Research that I have conducted, with the help of other VCU faculty and graduate students, corroborates these broader findings on the local scale (Pitt, Wang, et al., 2018).

For our project we developed a GIS model for a neighborhood-scale study area near Manassas. We found that the 1900 total buildings in the study area could easily support almost 25 MW of rooftop PV capacity, or about half of the state’s current distributed PV total.

A team of VCU engineers then developed a hypothetical model of the distribution grid in the study area, and ran simulations to evaluate the ability of the distribution grid to host increasing levels of distributed PV market penetration. Their model sought to optimally locate PV systems at places on the grid where they would benefit grid operations the most. They found that optimal PV placement could reduce system-wide energy losses and voltage deviations, while avoiding reverse power flows, at market penetrations up to at least 20%. In addition, their model demonstrated that placing distributed PV on commercial buildings is particularly valuable to grid performance, as the peak energy demand for those buildings occurs during the mid-afternoon, when solar PV production is at its highest point.

Long-Term Limitations of Net-Metering

The research findings outlined above indicate that, given the lack of discernable negative impacts from distributed PV at current or anticipated levels, Virginia should be in no hurry to alter the terms of its current net metering policies, other than to raise the 1% cap on net-metered renewable energy within each electric utility service area.

However, steps do need to be taken to prepare for the much higher levels of distributed PV that can and should occur in the longer-term. This includes technological solutions related to energy storage and smart grid technology, but we also need longer-term policy solutions that send appropriate price signals so that distributed PV growth occurs in the right way.

In the fall of 2017 I released a study that explored issues around net metering by interviewing individuals representing electric utilities, the solar energy industry, and other affected stakeholders (Pitt, 2017). The interview participants agreed that net-metering has long-term limitations, such as:

• It discourages customers from investing in energy storage, participating in demand response programs, and/or taking other steps to shift their net loads away from peak-demand periods.
• It discourages west-facing PV systems, which produce less electricity overall than south-facing systems, but reach their peak production later in the afternoon, closer to high-value peak demand times.
• It does not incentivize the installation of distributed PV systems in locations that would provide maximum benefits to the grid (i.e., through voltage regulation or other grid support services)
• It values electricity exported from PV systems differently based on the type of electricity customer that owns that system (i.e., residential customers are compensated at residential retail rates, while commercial customers are compensated at commercial retail rates).

In light of these limitations, I conducted case studies of “post-net-metering” models that had been proposed by stakeholder groups in Maine, Minnesota, and New York. All of these models drew from the results of a VOS study to identify distinct component values for the electricity produced by DPV systems. Each of those models would offer a new compensation rate based on a new value-of-solar study prepared for that state. Most importantly, these new models would seek to account for the unique values that solar can provide at different times and places.

Recommendations for the Virginia Energy Plan

I recommend that Virginia pursue a two-phased approach to the regulation of distributed solar PV. The goal of the first phase should be to rapidly expand distributed solar PV capacity throughout the Commonwealth, with the goal of meeting 5% of our total electricity demand with distributed PV by the year 2025, if not earlier. The following policy recommendations would help to achieve that goal:

• Increase the cap on distributed solar PV capacity within each electricity service area to 5% of peak demand. As noted above, the prevailing evidence from research across the country demonstrates that this threshold can be reached without incurring significant costs or causing major technical problems for the grid.
• Institute modest financial incentives, such as tax credits or rebates, to encourage further expansion of distributed PV capacity. Given the decreasing cost of solar technology, only a small financial incentive would be needed to make distributed solar investment cost-effective for the majority of Virginia’s residents and business owners.
• Adopt legislation that clearly establishes the legality of third-party solar energy Power Purchase Agreements throughout Virginia. The PPA model is essential for taking advantage of solar energy potential on buildings owned by schools, government agencies, and non-profits that cannot take advantage of the existing federal solar energy tax credit.
• Implement true community solar programs, which – unlike the proposed “community solar” tariff in the Dominion territory – would provide opportunities for Virginia residents and businesses to pool their resources and collectively purchase solar PV installations. Such models provide access to solar energy for individuals who otherwise are not able to own their own systems (e.g., renters and anyone else who does not own a building or property with good solar access), and they facilitate the development of medium-sized PV arrays in ideal locations.
• Enable Property Assessed Clean Energy (PACE) loan programs for the residential sector and encourage local jurisdictions to implement these programs.
• Work with local jurisdictions to implement streamlined local permitting processes for distributed PV installations.

For the second phase of my recommendations, Virginia should begin the process of developing a future “post-net-metering model” for compensating distributed PV system owners for the electricity that their systems offload to the electrical grid. This post-net-metering model should overcome the limitations of current net-metering arrangements and pave the way for a future electricity system in which large amounts of distributed PV are properly located on the grid and integrated with energy storage, demand management, and smart grid principles.

A key component of the post-net-metering model is that it should compensate each distributed PV owner based on the sum of the individual component values associated with their unique system. This would require the state to commission an independent value-of-solar study to quantify the various costs and benefits that distributed PV provides to the grid. 

Many of the component values that make up a post-net-metering tariff would be fixed for all systems within a given utility service area. However, it is important that the post-net-metering model also take into account the unique locational and temporal benefits associated with each system. This means that research, such as the VCU study described above, would be needed to identify locations on the grid where distributed PV systems can provide the most value. This would also require the widespread implementation of advanced electrical meters, so that PV owners could be compensated at a higher rate for electricity delivered at times of peak electricity demand, as well as time-of-use retail electricity rates that reflect that temporal variability in electricity costs and send those price signals to consumers.

Works Cited

Barbose, G. (2017). Putting the potential rate impacts of distributed solar into context. https://emp.lbl.gov/publications/putting-potential-rate-impacts

Denholm, P., & Margolis, R. (2008). Supply curves for rooftop solar PV-generated electricity for the United States. http://www.nrel.gov/docs/fy09osti/44073.pdf

Gagnon, P., et al. (2016). Rooftop solar photovoltaic technical potential in the United States: A detailed assessment. http://www.nrel.gov/docs/fy16osti/65298.pdf

Hallock, L., & Sargent, R. (2015).  Shining rewards: The value of rooftop solar power for consumers and society. https://environmentamerica.org/sites/environment/files/reports/EA_shiningrewards_print.pdf

Lopez, A., et al. (2012). U.S. renewable energy technical potentials. http://www.nrel.gov/docs/fy12osti/51946.pdf

Navigant Consulting (2016). Virginia Solar Pathways Project: Study 1: Distributed Solar Generation Integration and Best Practices Review. http://solarmarketpathways.org/wp-content/uploads/2017/09/DVP_DG-Transmission-and-Distribution-Grid-Integration-Study-1.pdf

Paidipati, J., et al. (2008). Rooftop photovoltaics market penetration scenarios. www.nrel.gov/docs/fy08osti/42306.pdf

Pitt, D. (2017). Evaluation of compromise post-net-metering models for distributed solar energy. https://cura.vcu.edu/media/cura/pdfs/cura-documents/Pitt-CompromisePost-Net-MeteringModelsforDistributedSolarEnergy-FINAL(1).pdf

Pitt, D., Wang, Z., et al. (2018). Optimizing the grid integration of distributed solar energy. https://wilder.vcu.edu/media/wilder/documents/OptimizingtheGridIntegrationofDistributedSolarEnergy.pdf

U.S. Department of Energy. (2012). SunShot vision study. www.energy.gov/sites/prod/files/2014/01/f7/47927.pdf 

U.S. Energy Information Administration. (2017). Electric Power Monthly with Data for December 2016. Table 6.2.B. Net Summer Capacity Using Primarily Renewable Energy Sources and by State. https://www.eia.gov/electricity/monthly/archive/february2017.pdf

U.S. Energy Information Administration. (2018). Electric Power Monthly with Data for December 2017. Table 6.2.B. Net Summer Capacity Using Primarily Renewable Energy Sources and by State. http://www.eia.gov/electricity/monthly/

Virginia State Corporation Commission (2011).  Response to net energy metering information request from the Virginia General Assembly. [Not online – contact author]