Reframing LCOE: New interactive tool expands the scope of generation cost comparisons

The tool from UT-Austin compares the full cost of electricity from ten generation resources in every US county

New research comparing the costs of generation resources has produced an interactive tool that analysts say gives some of the most granular insights available to the public today.

Nearly everything there is to know about electricity, from consumer and generator economics to the impacts of pollutants that come with its generation, is covered in The Full Cost of Electricity (FCe-) from the University of Texas (UT) at Austin.

“We wanted more detail about the cost to generate and deliver electricity because it goes far beyond what shows on the utility bill,” said Fred Beach, assistant director of the University of Texas at Austin Energy Institute.

A key white paper in the two-year, thirteen-study project details the Institute’s new online levelized cost of electricity (LCOE) calculator, Beach, a principal investigator on the study, told Utility Dive. Along with the typical cost factors in generating and delivering electricity, this calculator “includes locational information and costs for crucial externalities like health and environmental impacts not available with other interactive tools.”

LCOE is the cost metric most commonly used to compare generation technologies. It is the estimated cost “for a particular electricity generation plant to produce a kWh of electricity over its expected lifetime,” according to “New U.S. Power Costs,” the white paper that explains the LCOE tool.

The LCOE calculator’s detailed analysis was used to create energy resource maps from new levels of LCOE data granularity, according to Joshua Rhodes, a postdoctoral research fellow at the Institute. Rhodes was the lead researcher for the white paper and headed the team that built the LCOE tool.

“We calculated LCOE, county by county across the U.S., for ten types of new generation sources,” Rhodes told Utility Dive. Two generation types were coal-burning power plants — one with 30% carbon capture and sequestration (CCS) and the other with 90% CCS.  

Two types were natural gas combined cycle plants with and without 90% CCS. A fifth type was a new natural gas combustion turbine, or peaker plant. The other five types of generation were a nuclear facility, an onshore wind project, a utility-scale solar PV project, residential solar generation, and a concentrating solar power project with six hours of storage.

“The tool is interactive and users can change input assumptions, which allows it to be entirely useful in screening for where these technologies might be economic,” said Eric Hseih, finance and incentive analysis director at the U.S. Department of Energy (DOE).

But, he added, “you are not going to build a power plant based on the findings derived from this kind of broad-based online application.”

“After you know the LCOE, it takes a lot of analysis to get to the decision to build,” agreed Nate Blair, manager of the strategic energy analysis center at the National Renewable Energy Laboratory said. “But for comparing LCOEs at a location, this is a valuable and useful tool and the maps show key emerging energy sector themes like low cost natural gas and falling costs for wind and solar.”

Whitney Herndon, a research analyst at the Rhodium Group agreed, with a caveat.

“The good news is policymakers can take the information from this LCOE analysis and contrast it with what they see in the real world and use it to make good policy,” she told Utility Dive. But, she added, LCOE may overlook or generalize “risk profiling, reliability, and non-financial factors” crucial to the decision to build.

Input elements to the new UT LCOE estimates 


Old LCOE, meet the new LCOE

Traditional LCOE has become the standard for cost comparisons because it "offers several advantages as a cost metric,” the UT white paper notes. It can “normalize costs in a consistent format across decades and technology types.”

The LCOE calculation, including externalities, was a starting point for the project, Rhodes said. “When I realized location really mattered, we decided to pursue the spatial analysis.”

The UT LCOE framework “preserves the benefits of the conventional LCOE while adding in environmental externalities and geographic resolution,” the paper summarizes.

The paper provides several recognized equations for calculating traditional LCOE. The primary one includes the capital costs of the plant, fixed and variable operations and maintenance costs, and the average capacity factor over the plant’s lifetime. It also includes conventional plants’ heat rate and fuel price, which do not apply to wind and solar.

The tool adds costs for "harms done by greenhouse gas emissions and air pollutants and, a spatial scale, which allows the calculation of how prices, costs, and other factors differ from place to place,” Rhodes said.

Other LCOE calculations, like the widely-recognized Lazard Cost of Energy Analysis, give each type of generation a range of values, “but they don’t specify where the high and low prices are,” Rhodes added. “Location-specific values are important because some technologies are being deployed where the resource is not very good.”

For wind and solar, Rhodes analyzed 5 km by 5 km sites in every county of the continental U.S.. Because generation from coal, natural gas, and nuclear plants is dispatched, they were localized to the level of the 22 sub-regions defined by the North American Electric Reliability Corporation.

The solar and wind capacity factors necessary for the LCOE calculations were based on the site’s resource availability. Thermal plant capacity factors were based on the average capacity factor of a same-fuel plant in the subregion, Rhodes said.

The categories of input in the LCOE tool have default values but, because the tool is interactive, can be adjusted to match real-time prices and emissions costs, he pointed out. Interacting with the tool begins with the user identifying the county, the fuel source, the technology, and the technology’s capacity factor.

The tool populates values for the variable costs, which include the fuel cost and the fixed and variable operational & maintenance costs. It also populates values for fixed costs, including overnight capital costs, firming costs, the transmission distance and cost of transmission, and the mortgage period and interest rate.

Finally, it populates values for externalities at the location, including upstream and downstream CO2 emissions and SO2, PM25, NO2, and PM10 pollutants.

For Beach, the tool produced a critical insight.

“In the massive calculation of what should or should not be the energy choices for the future, there are hundreds of variables,” he said. “But the tool shows there are three first order energy variables that really make a difference in the LCOE.”

The price of natural gas, the capital costs of building wind and solar projects, and the price attributed to greenhouse gas emissions “make huge differences very quickly,” he said. In contrast, the tool reveals that “even if coal is free, the cost of the plant and the cost of the emissions are so high that coal is still an untenable choice.”

It also shows that nuclear power’s very high capacity factor makes it almost economically viable despite its very high upfront capital costs, he added. Its very low carbon intensity will likely make it cost competitive with wind and natural gas if the cost of emissions rises significantly, he added.

The minimum cost technology for each county, not including externalities

LCOE and the full cost of electricity

LCOE was the unifying factor in the FCe- project, according to Beach. Its many white papers examine how consumers, generation technologies, system hardware, and externalities not usually included in traditional cost evaluations interact in the electricity cost equation. They all come back to questions of cost.

One paper, “Impact of Renewable Generation on Operational Reserves Requirements,” investigates the costs needed to balance the variability of wind and solar and comes to a new conclusion, Beach said.

The paper analyzes costs incurred by the Electric Reliability Council of Texas (ERCOT) to manage variability as wind generation on its system went from 200 MW to more than 15 GW over ten years, he said. “It shows the cost of electricity went down and the cost of managing the grid went down.”

The ERCOT system’s two 2-GW nuclear facilities had already required the grid operator's planners to prepare for a worst case of a nuclear plant suddenly tripping offline.

“That is an order of magnitude more severe than 2 GW of wind slowly fading over a half hour,” Beach said. “Because of the nuclear plant threat, the grid support had already been built in and wind added no extra cost.”

Another paper, “Integrating Photovoltaic Generation,” compares the need for infrastructure upgrades caused by high levels of PV solar penetration on three standard but different residential distribution circuits, Beach said.

It finds that 100% PV penetration is possible for one type of circuit without any upgrade cost, he said. A second type of circuit will allow up to 30% penetration with no upgrade cost. But, to get to 100% on it, PV owners would have to add smart inverters. A third circuit type requires a $30 per transformer investment to get to 100% penetration.

“The point of the study is to show utilities which types of circuits are needed going forward to allow higher penetrations of solar without the necessity of infrastructure upgrade costs,” Beach said. “Where existing circuits make smart inverters necessary, they should be required by code. Where transformers will require upgrades, utility planners can anticipate the costs.”

The tool allows policymakers, utility regulators, utility leaders and the general public to see that “there is more to the cost of electricity than the utility bill and to understand the things that affect the full cost,” Beach said.

“The big takeaway is that this whole project is about education,” he said. “That is why the LCOE interactive tool is key.”

Decisions will be made over the coming decades involving “investments of billions or maybe even trillions of dollars in infrastructure,” he said. “These decisions should be made on more than what is cheapest now, because each decision is a commitment for 30 years and each includes risks as well as benefits.”

The tool is technology agnostic, he added. It shows the risk of natural gas price volatility but it also shows the risk of the embedded carbon costs in manufacturing a solar cell. “Solar PV has a fairly big CO2 footprint and as the price goes down, solar use grows,” Beach said. “But if the price of CO2 goes too high, some of solar’s cost advantage is lost.”

Minimum cost technology for each county, including externalities

Externalities and the full cost of electricity

The question of externalities is fundamental to any full cost of electricity discussion, Beach said.

Health and environmental costs, however, remain external to the price of electricity both on a customer’s utility bill and in a utility's calculation of the cost of a new power plant.“This is a disservice to the utility's regulators and to its customers because customers pay that price and bear those costs in alternate forms, like health care costs and taxes that fund renewables incentives," Beach said.

The UT project addresses them in its “EPA’s Valuation of Environmental Externalities from Electricity Production” white paper. Beach argued this should not be controversial because economists commonly recognize externalities.

“Over the last five decades, we have done a pretty good job of cleaning up this country’s polluted land and water,” he said. “This generation’s challenge is cleaning up the air and the cost of doing it should be in utility rates because power plants are the biggest source of U.S. emissions after transportation.”

In building the LCOE calculator, Rhodes discovered there is a locational component to the harms done by pollutants. “They vary depending on how dense the population is and how clean the air is,” he said. “That was the final insight informing the decision to make the tool so spatially granular.”

There may be debate about the tool’s assumptions and inputs but its consistent framework “can inform policymakers and inform their debate,” he said. And because it is interactive and transparent, it allows the debate's next step. “An alternative capital cost or cost of carbon can be plugged in and an LCOE based on that assumption comes out.”

Minimum cost technology for each county, including externalities and availability zones with a low price on all forms of CO2

The limits of LCOE

Rhodes acknowledged that UT’s LCOE is limited by the lack of temporal data. It leaves out real time wholesale market prices and dynamic local grid conditions like load, generation, and line congestion. He hopes to update the input data annually but that will not capture changing factors completely.

“If demand is flatter, there will be more base load plants built but if demand fluctuates with heating or air conditioning loads, a grid operator will want more fast ramping resources, he said. “Where demand is being met with rising renewables penetrations, loads are changing and the choice of generation technologies is being influenced by system needs.”

The LCOE calculator is a screening tool, Rhodes said. “It shows a region’s most likely technology choices but also demonstrates the choices are not the same everywhere and may not always be based on economics.”

DOE’s Hsieh found the tool’s interactivity, geographic specificity, and interactivity to be examples of Energy Department technical data being innovatively used to support the private sector. “Anyone who is thinking about energy resources nationally also needs to be aware of local factors and should find this tool useful,” he said.

By allowing users to change inputs, the tool can “move the conversation about energy technologies beyond dated assumptions, Hsieh added.

Hsieh used the tool to test assumptions in DOE’s just-updated 2016 Revolution Now energy sector profile and found the reported county-level cost declines accurately reflected in UT’s tool. But, he acknowledged, “the outputs of models like this depend on the assumptions and maintaining this tool’s input data on market conditions, especially fuel prices and technology costs, will be challenging,” he said.

A soon-to-be-available DOE tool will allow users to model the financial impacts of federal and state policy mechanism changes on generation technologies regionally, Hsieh said. This LCOE tool could potentially benefit from a similar capability to test “what happens when the dial is turned on various policy levers.”

NREL’s Blair agreed the tool’s locational specificity and the maps that display that specificity are key advances.

“The solar LCOE in a county near Seattle is going to be different than the solar LCOE in a county near Phoenix and the tool has default values for those differences,” he said. “Showing the story that way prevents the tendency to use a single national LCOE and throwing away the many local resource variations.”

The shortcoming of LCOE is that it lacks the precision of NREL’s complex, multifactorial models that calculate “the costs of grid integration and all the system factors involved,” Blair said. “But this tool is a good step in the direction of complexity without being burdensome.”

Blair, like Hsieh, sees much work in keeping the tool’s default values usefully updated with changing market prices. It is also important to recognize that “mandates, load shapes, the existing resource mix, and grid operations issues may play a greater role than LCOE in deciding between technologies.”

Rhodium’s Herndon sees LCOE as effectively comparing the costs for different types of resources. “This analysis in particular shows how what we assume about fuel prices, externalities, and other real world operators can change the bottom line.”

It also exposes the limits of LCOE as a decision-making tool, she observed. It finds that new nuclear is cost-effective “especially when externalities and availability constraints are considered,” she pointed out. LCOE does not recognize the challenges in a nuclear facility’s high upfront capital cost and in the protracted construction that delays investors’ returns.

On the other hand, accurate LCOE analysis can be the basis of remedying policy that leads to greater capital availability and fewer construction delays, Herndon said.

Filed Under: Generation Transmission & Distribution Solar & Renewables Regulation & Policy Technology