Revisiting the Electric Energy System : Leveraging Distributed Generation
The current architecture of the North American Electric System has evolved over 130 years. The current reliance on large thermal, central station power generation is an artifact of both technological innovation as well as environmental and economic forces which developed over much of the last century. The original model was much like the current Independent Power Producer (IPP) model in that a developer would identify an attractive thermal and electrical load and develop a generating station using project financing techniques. Over time, the power generation companies expanded to serve surrounding electrical loads and in several successive waves of business aggregations and roll-ups, larger regulated monopoly utilities grew. As society forced the generators out of the city limits for environmental reasons and utilities found it beneficial to interconnect their contiguous loads, our current model evolved. Higher pressure and temperature super critical boilers made central station economics attractive. High voltage materials science gave us the capability to traverse great distances with the large scale interconnected Transmission System to deliver the power to end use customers remote from the source of production. In the last decade, a shift to renewable generation resources has caused much discussion and concern regarding the role of the current “grid” and the utilities that run them. This paper suggests a potential future architecture which includes significant penetration of distributed renewable resources and a continuing reliance on the interconnected grid and the continuing role for the monopoly utility business model.
Where We Are
A resurgence of the renewable or alternative energy business has taken place over the last decade, fueled in large part by state and federal policy mandates, incentives and significant reductions in project costs. Despite the significant reduction in economic activity beginning with the recession of 2008, growth in the renewables sector accelerated particularly in the deployment of wind and solar PV[i].
What was once a niche has grown to become a business concern for utilities and regulators. The intermittency and low capacity factors of wind and solar PV relative to base loaded thermal power stations has raised operating concerns. Revenue erosion as customers use less energy and capacity of the grid or actually generate back into the grid creates concerns for the utility business model and non-participating customers are seen as de facto financial support for those installing renewable energy resources.
How We Got Here
Given the high degree of interconnection in the modern electric grids in the US, it is easy to forget that they did not start out that way. The advent of the modern electric utility business began with Edison’s Pearl Street Station in New York City in 1882. The steam engine driven dynamos were located in close proximity to the lighting load they served and that model was adopted and refined even as DC generation was replaced by Westinghouse’s Tesla designed AC generators—generation at the load or what we call today DG.
The early business model looked a lot like the Independent Power Production business model of today except that the generation was located in close proximity to the load it intended to serve. A new plant required a commercially viable host and relied upon project finance with long term off-take agreements for heat and power. Over time the generation reached out to serve surrounding loads. The system as we know it today grew organically, one project at a time until the first wave of industry aggregation. As utilities grew through organic load growth and merger and acquisitions of adjacent systems they benefitted from interconnection with neighboring units and/or utilities.
Over the early part of the 20th Century technological innovation in electric energy generation, transmission and distribution facilitated the location of larger fossil fueled plants remote from the load centers. The location of these newer plants was driven to larger bodies of cooling water or nearer to their sources of fuel and so the economics favored the evolution of large super critical generating units which of necessity had to be connected to longer and higher voltage transmission lines serving geographically separated load centers. According to US EIA data, average energy consumption between 1950 and 1973 grew at an average 5 to 7% Compound Annual Growth Rate[ii]. At that rate, new capacity requirements doubled about every decade. In the post-oil embargo era, energy consumption dropped precipitously to 1.5% CAGR on average. Units committed to serve the 7% CAGR growth curve, including many over-cost nuclear units, soon became excess capacity that fueled the State and Federal policy moves to vertically disaggregate utilities in the 1980’s and 90’s. The effects of PURPA and subsequent deregulation of vertically aggregated utilities marked a return to the IPP model for new generation capacity in most of the United States.
Where We Are Going
According to the DESIRE database[iii], since 1983, 38 states and the District of Columbia have adopted Renewable Portfolio Standards, 30 of which have mandatory requirements and 8 of which have voluntary compliance requirements. In addition, 44 States including DC have established Net Energy Metering (NEM) requirements. The RPS requirements are a patchwork of policies that vary from jurisdiction to jurisdiction. Some include all forms of renewables, and 10 of them call for set aside requirements for solar distributed generation.
During this same period of time, Utility Deregulation or Restructuring has occurred in 16 states plus DC vertically disaggregating the Utilities in those jurisdictions.[iv] To further complicate the picture 34 states and DC partially or totally participate in organized wholesale markets for electric power transactions through a Regional Transmission Operator (RTO) or Independent System Operator (ISO)[v]. The upshot of these changes in policy and regulation has been a return to reliance on private investment for new or replacement generation assets in most, but not all of the country.
However, uncertainty over future carbon regulations has all but killed coal fired base load construction. According to the National Electric Technology Lab report Tracking New Coal-Fired Power Plants, there are less than 1 GW of new coal capacity scheduled in 2013 and 2014 and virtually none out to 2020 after that compared to planned retirements of coal plants on the order of 5-8 GW per year.[vi] Expected capacity will come from natural gas and renewable resources. In fact, just under half of all new generating capacity added in the US last year was classified as renewable. The bulk of the renewable additions last year were grid connected wind. However, 2013 projections are for wind additions to begin to fall as solar additions grow and the bulk of those contributions to be distributed solar PV according to the Solar Energy Industry Association (SEIA) report titled U.S. Solar Market Insights: Q1 2013.[vii]
Most of the focus by developers through the 2000’s has been on large scale on-shore wind and both grid connected as well as distributed solar PV. But, grid connected renewables are just a new version of central station power production. The problems attendant to central station power sources affect the siting and permitting of these projects as well including long distance transport of the energy to the load centers. As Thomas Edison knew, and we are rediscovering, there are substantial benefits to be derived by placing new renewable sources closer to the load they serve.
A New, Old Model
Solar PV penetration has grown from virtually nothing in 2000 to about 4 GW at the end of 2011[viii]. An additional 3.3 GW of capacity was installed by the end of 2012 and the projection by SEIA/GTM Research predicts an additional 4.4 GW of solar PV in calendar year 2013[ix]. While much of the previous addition has been large scale grid connected projects, the tide has turned toward smaller, distributed solar PV which is expected to make up the bulk of the additions in coming years.
Clearly, the pace of deployment has been driven by the various state and federal incentives available to developers and owners. However, the continued growth has most recently been driven by solar PV cost reductions. Since 2007, PV panel prices have fallen over 70%. Solar panels can be had in volume at about 70 cents/Watt delivered and continued price pressures are expected into the future.
Total installed costs range in the $2.50-4.00/Watt range and not much work has yet been done in reducing the balance of systems or financing costs for PV systems[x]. However, at current developed costs, assuming a 30 year life and a conservative estimate of future O&M, the US Energy Information Administration calculates the Levelized Cost of Energy (LCOE) to be on the order of 13 cents/kWh[xi]. Actual developments I have experience with have come in around 8 cents/kWh all-in, making the LCOE comparable to pre-recession wholesale power pricing. Even at the higher, more conservative estimate of US EIA though, that LCOE is below current retail rates in high cost states like New Jersey, New York and California. So the question must be asked, what is holding back even more rapid deployment of distributed solar generation particularly in high cost of service areas of the country?
The barriers are numerous but the most significant, in my experience are the balkanization of markets, uncertainty of policy with regard to the future of incentives, an over-reliance on tax-equity financing and project finance and a general lack of understanding of the value that distributed generation resources brings to the grid.
The balkanization of markets is the artifact of past policy decisions by both state and federal governments taken at different points in time without an overarching strategy, intended to solve then current problems locally or regionally. There is not likely to be any immediate change in the status quo given that the Congress is divided on the issue of a national energy strategy. Absent a crisis, we cannot expect any political change in the near term.
Many of the projects sited in the distribution system, of necessity, will be below 1MW of capacity limited by both space and customer load as well as distribution system design. To date, the prevailing wisdom has been that projects at this size and below cannot scale because the project finance model creates high cost in both acquisition of capital and legal costs making development economically unattractive. Given the relatively low penetration rates so far and the raging debate over new financing tools, we have to conclude that there is merit to this argument.
Those familiar with electrical design and construction of these projects point out that there can be economies of both scale and scope created once a substantial pipeline of projects is established. There are economies that come from both the supply chain as well as learning’s that are achieved from replication and most of those economies are yet to be achieved in this market. Achievement of these benefits though cannot be realized without a better funding vehicle than is currently in place.
Tax equity financing has undoubtedly been an essential tool in the evolution to date. The explosion in solar deployments in 2010-2012 was, at least in part the result of the conversion of that incentive into a cash payment during the economic stimulus funding which has now expired. The limited availability of tax-equity investors and the high cost of capital they command will not allow the current trajectory to continue.
The rapid growth of the electric utility industry in the post-World War II era relied upon the ability of utilities to securitize their capital additions. Access to stable, predictable, long term sources of funding can have a similar impact on the emergence of the distributed generation market. But the ability to finance future projects with bond proceeds is not available to all but a very few large developers. Again, utilities are in a position to access those capital markets at a cost and tenor that can make the deployment of distributed generation grow. After all, these resources are long lived, utility like assets with proven predictable operating characteristics and lives.
Most customers do not want to be generation owners and operators. We know this from utility history as large commercial and industrial consumers sold off or closed down their own generation sources in favor of utility supply during the mid-1900’s. The capital costs of upgrades and replacements coupled with future fuel price volatility convinced many of them that the utility option made the most sense. Current history in the solar renewable market place suggests that most hosts prefer to acquire solar energy through a Power Purchase Agreement wherein they get stable predictable long term pricing for electricity without the burdens of upfront capital or future O&M costs. Again, this suggests that those familiar with owning and operating long lived assets are best positioned to take advantage of this market development phase of growth.
In order for this suggested model to work, two critical ingredients are required. First, the utility must be convinced of the value proposition in moving to a more distributed generation platform. Secondly, the customer must be convinced in the long term value proposition in allowing the distributed generation on their site/building.
Much of the current policy debate has focused around elimination of incentives including Net Energy Metering driven largely by the concern over utility revenue erosion as market uptake accelerates. The fact of the matter is that we are going to see the evolution of renewable distributed resources in the interconnected grid. The only questions are going to be who wins and who loses? I posit that there need not be any losers.
It has been suggested by some that the intermittent nature of solar distributed generation will cause operational issues in the interconnected grid at even modest penetration levels. The National Renewable Energy Laboratory (NREL) report Renewable Electricity Futures Study concluded that: “The central conclusion of the analysis is that renewable electricity generation from technologies that are commercially available today, in combination with a more flexible electric system, is more than adequate to supply 80% of total U.S. electricity generation in 2050 while meeting electricity demand on an hourly basis in every region of the United States.”[xii]
Will there be changes and adaptations required? Certainly. But the challenges can be met and do not require unknown or unavailable technology for success. What about customer objections to utility owned and operated assets? My 33 years of utility operational and executive management experience suggest that customer dissatisfaction stems from two main sources. They are price and availability. Customer dissatisfaction over either or both of these issues generally translates into a call for “competition” or “control”. The calls for competition in energy production during the rate shock years of the 1970’s led to PURPA and partial deregulation as mentioned before. During those years and in more recent time’s customer aggregation and municipalization were fuelled by concerns over rapidly increasing rates and uncertainty as to future costs by large industrial and commercial customers.
Availability in the utility industry is generally measured as an average for all customers over a year based on 8760 hours/year. According to a 2012 report from Lawrence Berkeley National Laboratory, the average outage duration and average frequency of outages for US customers is increasing at a rate of approximately 2%/year.[xiii] A system designed to achieve 99.97% availability is seen as insufficient given the digital age we live in. While the report is not conclusive as to the cause of the trend, those in the utility industry know that customer perception of reduced reliability/availability has translated into customer and regulatory pressure for change.
That change, in some minds, requires a massive new investment in capacity which of course works against the primary concern regarding price. By some estimates, upwards of $1.5 Trillion are needed to bring the system to a level of performance that would satisfy customer demands. What if, instead of spending money to reinforce the Transmission and Distribution Systems we utilized distributed generation resources and semi-autonomous “smart grid” architecture to isolate an outage cause to the smallest possible set of affected customers? There is likely to always be multiple single points of failure in the T&D system. Cars careen out of control and strike poles, contractors dig up underground cables, insulators get contaminated and arc over, customer loads or faults can cause equipment failures upstream. The ideal is not to try to eliminate them but to limit their effect to the smallest group of customers, make their occurrence detectable, rapidly by-passable and easily repaired and returned to service.
The Rocky Mountain Institute’s Electricity Innovation Lab issued a report summarizing the known body of work on the benefits and cost studies of solar PV.[xiv] In all, 15 cost/benefit studies undertaken by regulatory bodies, electric utilities, national labs or other organizations were reviewed and assessed as to the range of cost and benefits and study methodology. Differences in methodology and assumptions make direct comparisons of the studies not useful.
However, there were a number of observations that support a view that benefits are net positive. There is general agreement that energy cost reduction due to less energy production, energy loss reduction and avoided capacity charges are possible and fairly easily quantified. Likewise it is understood that distributed generation can provide ancillary services to the transmission and distribution system by way of reactive supply, voltage regulation, frequency response and congestion reduction though there is no general agreement as to the value of such services. Other posited benefits such as elimination of criteria pollutants, water and land use, carbon reduction, jobs and economic growth, national energy security, enhanced reliability and/or resiliency, and fuel price hedging are still being debated and are not easily monetized.
The net value from the NREL study, which is typically in the middle of the pack among the studies, amounted to about 28 cents per kWh for distributed PV. Of course there continues to be much debate as to the proper methodologies and assumptions to use in making these calculations but it should be clear that there is sufficient benefit from distributed generation resources that, particularly given that solar PV is competitive with new sources of generation, we simply cannot ignore it.
As the various studies acknowledge, it is going to be difficult to unbundle all of these points of value in order to make a truly fair comparison. But, we do not need to if we assume a utility ownership model. The fact is that the intrinsic value will be released when the utility makes the decision to install distributed generation instead of installing a fossil fueled central station generator or a new Transmission line to eliminate or reduce congestion. Assuming a utility must make an investment in new G,T or D capacity to serve its load growth or replace retiring facilities, the customers are benefited when that decision is displaced by distributed generation resources. Of course, this statement assumes significant market penetration of distributed resources are achieved.
Importantly though, this model can be economically attractive in the current grid architecture. The adoption of widely distributed generation resources is an essential, but not sufficient condition for the establishment of the grid of the future. The addition of “smart” inverters and distributed energy storage holds promise to take the grid to its ultimate performance. Smart inverters, capable of autonomous or semi-autonomous operation already have the ability to provide Volt/Var regulation and islanding capabilities in the event of loss of the grid for outage events. Coupling this capability with distributed energy storage technology, whether collocated with the distributed solar PV or not, effectively makes those resources “dispatchable”. The distributed energy storage can be used to mitigate or eliminate any intermittency, provide frequency regulation and demand response services, and load shaping to maximize the value of the solar resources.
Finally, in combination distributed solar PV and distributed energy storage coupled with smart grid enabled equipment can provide for significant reduction in both customer outage frequency and duration. Essentially achieving the long sought after “self-healing” grid.
Achievement of these benefits is by no means assured. The single largest impediment to releasing the value required in order to make the investments worthwhile rely on the development of a business model that will attract reasonably priced capital to accelerate adoption and penetration to the level where these benefits can be achieved.
I argue that model already exists. Utility ownership of these distributed resources obviates the need to eliminate balkanized policies, provide new or renewed financial incentives or provide new tools to access capital markets all the while ensuring rapid deployment and long term stable ownership and management of the assets for the benefit of the utility, its customers and ultimately the nation.
This is not a call for utilities to become large scale EPC contractors or to insource the development process. Just as utilities do not design or construct large central station generation and contract out a lot of the T&D design and construction work, they can rely on the already available pool of developers and EPC contractors for the essential work. Rather, this is a call for them, in exchange for all of the benefits enumerated, to bring their financing muscle to the task of funding the build out and to utilize their expertise in the integration of these assets into the grid of the future.
No doubt there are those skeptical of such a model but it is relevant to point out that over the last 130 years, the electric utility industry has created one of the greatest machines known to man in the form of the interconnected grid. Access to readily available and cheap energy has been the engine of the US economy for at least the last 50 years. A look at US EIA data over that period reveals that the “real” price of electricity delivered at the end of 2010 was the same as it was in 1960! Over that period of time, electricity prices were effectively indexed to GDP growth and the data are irrefutable that productivity improvements brought about through electrification of work were the reason. There is every reason to believe that the next 100 years can achieve at least the same benefit, all the while securing the nation’s economic, energy and environmental security.
Utilizing a Rate Base, Rate of Return approach eliminates many of the barriers presently preventing rapid scaling of renewable distributed resources. Providing the utility a reasonable return of and earnings on its investment over the asset life in return for using its balance sheet to access cheap capital markets is the model that built the current grid and provided the energy that fueled our economic growth. Utilization of this model to build renewable generation and eventually storage technologies will allow us to unlock the value of distributed energy resources in a way that benefits all constituencies- Utility, Customer, Regulator-through stable and affordable energy supply, a more resilient and capable grid for this next century, and a safer, cleaner and more secure world.
[i] 2011 Renewable Energy Data Book, US DOE,Energy Efficiency & Renewable Energy (EERE), http://www.nrel.gov/docs/fy13osti/54909.pdf
[ii] Annual Energy Outlook 2013, Market trends-Electricity, US EIA, www.eia.gov/forecasts/aeo/MT_electricity
[iii] DESIRE Database, operated by the North Carolina Solar Center at N.C. State University, http://www.desireusa.org/rpsdata
[vi] Tracking New Coal-Fired Power Plants January 13, 2012, National Electric Technology Lab, http://www.netl.doe.gov
[vii] Solar Energy Industries Association/GTM Research Solar Industry Data Q1 2013 Report, http://www.seia.org/research-resources
[viii] Interstate Renewable Energy Council, 2012 Update & Trends Annual Report, http://www.irecusa./publications
[ix] Solar Energy Industries Association/GTM Research Solar Industry Data Q1 2013 Report, www.SEIA.org
[x] News Release US DOE Lawrence Berkeley National Lab, newscenter.lbl.gov/news-releases/2013/08/12/installed-price-of-solar-photovoltaics
[xi] U.S. Energy Information Administration, Levelized Cost of New Generation Resources in the Annual Energy Outlook 2013, http://www.eia.gov/topics/analysis&projections
[xii] Renewable Electricity Futures Study (Entire Report) National Renewable Energy Laboratory. (2012). Renewable Electricity Futures Study. Hand, M.M.; Baldwin, S.; DeMeo, E.; Reilly, J.M.; Mai, T.; Arent, D.; Porro, G.; Meshek, M.; Sandor, D. eds. 4 vols. NREL/TP-6A20-52409. Golden, CO: National Renewable Energy Laboratory. http://www.nrel.gov/analysis/re_futures
[xiii] Joseph Eto, et al., An Examination of Temporal Trends in Electricity Reliability Based on Reports from U.S. Electric Utilities, Lawrence Berkeley National Laboratory
[xiv] A REVIEW OF SOLAR PV BENEFIT & COST STUDIES, Rocky Mountain Institute, Electricity Innovation Lab, ww.rmi.org/elab_emPower