What Is LCOE? How to Calculate Levelized Cost of Electricity for Frontier Power Projects

Every power project in a frontier market eventually reaches the same room. The mine’s general manager wants uptime. The finance director wants a number he can put in a board paper. The lender wants a metric that survives due diligence. The question that decides the project is always the same: what is LCOE for this plant, and can you defend the number line by line?

Levelized cost of electricity, or LCOE, is the most powerful number in power project economics and the most frequently misused. It compresses capital cost, fuel, operations and maintenance, financing and plant output into one figure expressed in dollars per megawatt hour. Used properly, it tells a mining company whether to build heavy fuel oil generation or a solar hybrid. Used carelessly, it hides the two variables that actually decide whether a frontier project succeeds: the cost of capital and the plant’s real capacity factor.

According to the International Energy Agency, around 600 million people in sub-Saharan Africa, roughly 47 percent of the population, still lacked access to electricity in 2024. Closing that gap will be decided by thousands of individual investment decisions, each one resting on an LCOE calculation. USP&E Global has built more than 150 power projects across more than 35 countries over 25 years, and our 350+ engineers have watched LCOE models succeed and fail in the field. Your project is the hero here. This guide is the map.

What Is LCOE? The Definition, the Formula and What It Actually Measures

LCOE is the average cost of producing one unit of electricity across the entire life of a power plant, expressed in present value terms. In plain language, it is the constant price per megawatt hour that a plant would need to receive to exactly break even over its operating life, once every dollar of capital, fuel, maintenance and financing has been discounted back to today.

The formula is a ratio of two discounted sums:

LCOE = (sum of all discounted lifetime costs) / (sum of all discounted lifetime electricity output)

Written out, the numerator is the sum, across every year of the project, of investment cost plus operations and maintenance cost plus fuel cost, each divided by (1 + r) raised to the power of that year. The denominator is the sum of net electricity generated in each year, discounted by the same factor. The term “r” is the discount rate, normally the project’s weighted average cost of capital, or WACC.

Three definitions matter before going further. CapEx is capital expenditure, the one-time cost of engineering, equipment, civil works and commissioning. OpEx is operating expenditure, the recurring cost of fuel, spares, lubricants, labour and overhauls. EPC means engineer, procure, construct, the contract under which a single contractor delivers a complete plant. O&M means operations and maintenance, the contract under which that plant is run.

Two authoritative bodies publish the reference data most lenders will benchmark your model against. The International Renewable Energy Agency publishes global weighted average LCOE by technology each year, and the International Energy Agency publishes cost of capital data for emerging and developing economies. Any LCOE model presented to a development finance institution should be able to explain why it differs from those benchmarks.

How to Calculate Levelized Cost of Electricity: A Worked 20 MW Example

The levelized cost of electricity calculation becomes concrete very quickly once real numbers are attached. The following worked example uses a 20 MW medium speed heavy fuel oil plant serving an island grid at a West African mine site, a configuration USP&E has delivered repeatedly. All figures are indicative engineering ranges for illustration, not a quotation.

1. Fix the capacity and the output. Net capacity is 20 MW. At an 85 percent capacity factor, annual net generation is 20,000 kW multiplied by 8,760 hours multiplied by 0.85, which equals 148,920 MWh per year.
2. Fix the installed CapEx. At an all-in installed cost of USD 1,200 per kW, including balance of plant, fuel farm, switchgear and civils, total CapEx is USD 24 million.
3. Annualise the CapEx using the capital recovery factor. At a 12 percent WACC over a 20 year life, the capital recovery factor is 0.1339. Annualised capital cost is USD 3.21 million.
4. Calculate fuel cost per MWh. At a specific fuel consumption of 0.210 kg per kWh and a delivered HFO price of USD 600 per tonne, fuel cost is USD 126.00 per MWh. This single line is the reason heat rate discipline matters more than CapEx haggling.
5. Add fixed and variable O&M. Fixed O&M of USD 1.6 million per year equals USD 10.74 per MWh at this output. Variable O&M, covering spares, lubricants and the overhaul reserve, is USD 14.00 per MWh.
6. Divide and sum. Total LCOE is USD 172.32 per MWh, or USD 0.172 per kWh.

The lesson is immediate. Fuel represents roughly three quarters of the levelized cost of electricity for a frontier thermal plant. A one percent improvement in specific fuel consumption is worth more than a five percent discount on the engines. That is why USP&E’s power plant engineering teams model heat rate at site ambient conditions rather than at ISO reference conditions, and why our operations and maintenance contracts are structured around fuel consumption guarantees rather than headcount alone.

Why Frontier Market LCOE Behaves Differently: Cost of Capital, Fuel and Capacity Factor

Frontier market LCOE is not a scaled version of European LCOE. It is a different shape, because the weight sits in different places. Three drivers explain almost all of the divergence.

1. The cost of capital is two to three times higher. The IEA reports that the cost of capital for energy projects in emerging and developing economies runs at least two to three times higher than in advanced economies or China. Its Cost of Capital Observatory recorded median weighted average cost of capital for utility scale solar in 2024 at 9.4 percent in Indonesia, 9.0 percent in Viet Nam and 8.0 percent in the Philippines, against a range of 5.0 to 6.5 percent in advanced economies. Investors identify political, currency, regulatory, off-taker and transmission risk as the five constraints to address first.
2. Financing now explains more of the cost gap than technology does. IRENA’s 2025 cost analysis found that country level macroeconomic conditions explain roughly 56 percent of the variation in financing costs, against 24 percent attributable to technology, a ratio of about 2.3 to one. IRENA also documented that global solar PV installed costs fell around 6 percent in 2025, yet solar LCOE stayed flat at USD 44 per MWh, because a higher cost of capital cancelled out cheaper equipment. Where a project is built now matters more than what is built.
3. Capital intensive and fuel intensive plants respond to WACC in opposite ways. This is the insight most models miss, and it changes technology selection. The table below runs the same 20 MW HFO plant from the previous section against an equivalent 20 MW solar PV plant at a 20 percent capacity factor and USD 1,000 per kW installed, across four discount rates.

Moving from an 8 percent to a 20 percent WACC raises the HFO plant’s LCOE by 10 percent. It raises the solar plant’s LCOE by 91 percent. A capital intensive, zero fuel technology is punished brutally by frontier financing costs, while a fuel intensive technology is comparatively insulated. This is why a solar hybrid that looks unbeatable in a Frankfurt spreadsheet can lose to thermal generation in Bamako, and why concessional or blended finance changes the answer entirely. The African Development Bank and the International Finance Corporation both publish instruments designed to compress exactly this gap.

Capacity factor is the third lever. Holding the 12 percent WACC constant, the same HFO plant delivers an LCOE of USD 172.32 per MWh at an 85 percent capacity factor, USD 185.79 at 60 percent, and USD 208.68 at 40 percent. Every point of availability lost to a failed turbocharger or a fuel supply interruption is inflation in your cost per kWh. Regional generation and consumption baselines for calibrating these assumptions are published in the World Bank energy data portal.

LCOE Benchmarks by Technology for Frontier Power Projects

IRENA’s Renewable Power Generation Costs in 2025, published in 2026, is the reference point lenders use for renewable benchmarks. Its global weighted average figures are set out below alongside installed cost and capacity factor, because an LCOE figure without those two inputs cannot be interrogated.

Since 2010, IRENA records solar PV LCOE falling 89 percent, onshore wind 71 percent and offshore wind 63 percent. In 2025, more than 90 percent of newly commissioned utility scale renewable capacity delivered power below the cost of the cheapest new fossil fuel alternative. The thermal side moved the other way: a gas turbine shortage driven partly by data centre demand roughly doubled the capital expenditure for a new combined cycle plant, reaching around USD 2,400 per kW in the United States, pushing gas fired LCOE towards USD 100 per MWh in high gas price markets while remaining in the USD 50 to 60 range where gas is cheap.

Two qualifications matter for frontier projects. These are grid connected global averages, not remote island grid figures, and they exclude the firming cost of intermittency. IRENA’s own firm LCOE metric, which prices continuous supply from solar plus battery systems at 95 percent reliability, fell below USD 85 per MWh in 2025 at high quality sites, down from above USD 100 per MWh in 2020, with four hour utility scale battery installed costs at roughly USD 140 per kWh. That firm figure, not the bare solar figure, is the one that competes with a diesel or HFO plant on a mine site.

Fuel Type Comparison for Frontier Power Projects

Diesel gensets carry the lowest CapEx and the highest LCOE. HFO reverses that trade, which is why it dominates West African mining, and why USP&E maintains dedicated HFO generator inventory. Mobile gas turbines such as the GE TM2500 win where the cost of not having power exceeds the cost of a higher heat rate. Hybrid power systems rarely replace thermal generation in frontier markets, but they routinely cut its fuel bill.

Five LCOE Mistakes That Sink Frontier Project Economics

1. Pricing the prime mover instead of the plant. Engine or turbine cost is typically a fraction of the installed price. Shipping, balance of plant, civils, grid interconnection and fuel reticulation frequently add one to two times the cost of the prime mover. An LCOE model built on an equipment quotation understates capital cost by a wide margin.
2. Using nameplate capacity as the output denominator. LCOE divides by net electricity delivered, not gross nameplate. Parasitic auxiliary load, derating at high ambient temperature and altitude, forced outages and planned maintenance all reduce the denominator. A plant modelled at nameplate will always show a flattering LCOE.
3. Applying the wrong discount rate. A 7 percent WACC borrowed from a European precedent applied to a Mali or Liberia project will make a solar hybrid look decisively cheaper than it is. Model the discount rate that your actual capital structure will bear, then test the answer across a range.
4. Omitting the overhaul reserve. Medium speed HFO engines require top end and major overhauls on defined running hour intervals. A model carrying only routine maintenance is not a levelized cost model, it is a short term forecast.
5. Comparing an intermittent LCOE against a dispatchable LCOE. A solar figure of USD 44 per MWh and an HFO figure of USD 172 per MWh are not comparable numbers. One delivers power when the sun shines, the other delivers it when the mill is running. Compare firm, dispatchable cost against firm, dispatchable cost, or compare fuel displaced per year.

Case Studies: LCOE Discipline in Practice Across Frontier Markets

Sierra Leone, London Mining, 16 MW HFO, EPC. USP&E was engaged to design and deliver a 16 MW heavy fuel oil power station for the Marampa iron ore mine near Lunsar. The plant used ten HHI 9L21/32 engines operating on RME180 fuel, and was completed ahead of schedule and within budget. Fuel selection was the LCOE decision: at Marampa’s load profile, HFO’s higher CapEx and longer build programme were repaid by a materially lower cost per kWh than the diesel alternative across the mine life.

Mali, Loulo-Gounkoto complex, 80+ MW rebuild plus 24 MW new capacity, O&M. USP&E performed a complete rebuild of more than 80 MW of Caterpillar engines at the Loulo-Gounkoto complex in western Mali and installed 24 MW of new zero hour Cat 12CM32 engines. On a related Caterpillar O&M contract, USP&E’s engine repair and refurbishment programme achieved a 66 percent reduction in average cost per engine against the previous contractor, USD 312,284 versus USD 928,618. Cost per engine is not LCOE, but it flows directly into the variable O&M line, and a reduction of that magnitude moves cost per kWh.

Madagascar, Molo graphite mine, thermal plus solar hybrid, EPC. USP&E acted as EPC contractor on the Phase 1 thermal and solar hybrid plant, combining generation with solar and battery energy storage, designed to reduce carbon dioxide emissions by more than 11,300 tonnes per year. The hybrid case at Molo was not a lower headline LCOE, it was fuel displacement on an existing thermal base, which is where hybrid economics usually work in frontier markets.

Across 25 years, USP&E has delivered more than USD 250 million in fuel and operating savings to clients, without a single lawsuit. Full project details are set out on the USP&E project experience and client references pages, including the Siemens gas turbine procurement case study where careful equipment sourcing removed USD 10 million from a client’s capital cost.

How to Select an EPC and O&M Partner Who Can Defend Your LCOE: Nine Criteria

6. Ask for the LCOE model, not the LCOE number. A credible partner hands over the spreadsheet with every assumption visible. If the discount rate, capacity factor and fuel price are hard coded and unexplained, the number is marketing.
7. Test heat rate at site conditions. Ambient temperature, altitude and humidity all degrade output and efficiency. Insist on performance figures at your site’s design conditions, not at ISO reference conditions.
8. Confirm the balance of plant scope in writing. Every site is unique in its grid interconnection, fuel storage and civil requirements. No used power station is ever complete. Ask specifically what is excluded.
9. Interrogate the overhaul assumptions. Request the running hour intervals, the parts scope at each interval and the reserve per MWh. This is the single most commonly understated line in frontier O&M budgets.
10. Check whether the same firm can operate what it builds. An EPC contractor with no O&M business has limited incentive to design for low lifetime cost. Availability guarantees only mean something when the guarantor runs the plant.
11. Verify in-country execution capability. Boots on the ground shorten outage response times, and outage duration drives capacity factor, which drives LCOE. Ask about local headcount, not regional headcount.
12. Review the fuel logistics chain. Delivered fuel price, not port price, belongs in the model. Ask who bears demurrage, inland haulage and quality risk, and require third party fuel analysis.
13. Confirm compliance credentials. ISO 9001:2015 and ISO 45001:2018 certification, FCPA and OFAC compliance, and a clean litigation record all reduce the risk premium a lender will apply to your project, which lowers your WACC and therefore your LCOE.
14. Ask for remote monitoring and data. Availability, heat rate and fuel consumption need to be measured continuously, not reconciled annually. Digital O&M platforms such as SmartPower convert LCOE from a pre-investment estimate into a managed operating metric.

Frequently Asked Questions: What Is LCOE and How Do You Calculate It?

What is LCOE in power generation?

LCOE, or levelized cost of electricity, is the average cost of generating one megawatt hour of electricity across a power plant’s full operating life, stated in present value terms. It is calculated by dividing the discounted sum of all lifetime costs, capital plus fuel plus operations and maintenance, by the discounted sum of all lifetime net electricity output. It is the standard metric for comparing generation technologies on a consistent basis and the metric lenders use to test project viability.

What is the LCOE formula?

The LCOE formula divides discounted lifetime costs by discounted lifetime output. In each year, investment cost plus O&M cost plus fuel cost is divided by (1 + r) to the power of that year, and those values are summed. The same discounting is applied to annual net electricity generation and summed. Dividing the first total by the second gives LCOE. The variable “r” is the discount rate, normally the project’s weighted average cost of capital.

What is a good LCOE for a power plant in Africa?

There is no single good number, because LCOE depends on fuel, capacity factor and cost of capital. Utility scale solar PV globally averaged USD 44 per MWh in 2025 according to IRENA, while a remote mining HFO plant will typically land between USD 150 and USD 220 per MWh, and diesel standby generation higher still. What matters is whether your LCOE beats the alternative available at your specific site, at your load profile, under your capital structure.

How does the discount rate affect LCOE?

The discount rate affects capital intensive technologies far more than fuel intensive ones. Raising the weighted average cost of capital from 8 percent to 20 percent increases a 20 MW HFO plant’s LCOE by roughly 10 percent, but increases an equivalent solar PV plant’s LCOE by roughly 91 percent. IRENA reports that country level macroeconomic conditions explain about 56 percent of the variation in global financing costs, roughly 2.3 times the share attributable to technology choice.

What is the difference between LCOE and cost per MW?

Cost per MW is a capital cost measure, the installed price of building capacity, expressed in dollars per kilowatt or per megawatt. LCOE is a lifetime cost measure, expressed in dollars per megawatt hour of energy actually delivered. A plant can have a low cost per MW and a high LCOE, which is exactly the case for diesel gensets. Capital procurement decisions made on cost per MW alone routinely produce the highest lifetime cost outcome.

Does LCOE include grid connection and transmission costs?

Standard LCOE typically includes the plant’s own interconnection equipment but excludes wider transmission network reinforcement, system balancing and backup capacity. That exclusion is why intermittent and dispatchable LCOE figures are not directly comparable. For frontier island grid and captive mining projects, the connection scope is usually inside the EPC boundary and therefore inside the LCOE, which is one reason those figures sit above published grid connected averages.

How long does it take to produce a bankable LCOE model?

A conceptual feasibility study producing an indicative LCOE typically takes 30 to 60 days and includes a site visit and data collection. A full bankable feasibility study, with geotechnical survey, grid interconnection study, load study and detailed engineering inputs, typically takes 60 to 120 days. Anyone offering a firm lump sum turnkey price and a defensible LCOE without either study is quoting a number, not modelling one.

Summary: Key Takeaways for LCOE Decision Makers

• What is LCOE: the discounted lifetime cost of a power plant divided by its discounted lifetime electricity output, expressed in dollars per megawatt hour, and the standard basis for comparing generation options.
• Fuel dominates frontier thermal LCOE. In the worked 20 MW HFO example, fuel accounts for 73 percent of cost per MWh. Heat rate discipline outperforms CapEx negotiation.
• Cost of capital dominates renewable LCOE. Moving from 8 to 20 percent WACC raises solar LCOE by 91 percent and HFO LCOE by only 10 percent. Frontier financing costs reshape technology selection.
• Capacity factor is the hidden multiplier. The same plant costs USD 172 per MWh at 85 percent availability and USD 209 per MWh at 40 percent. Availability is an economic variable, not just an engineering one.
• Benchmarks are a starting point, not an answer. IRENA’s 2025 global averages of USD 33 per MWh for onshore wind and USD 44 per MWh for solar PV are grid connected figures that exclude firming, and Africa’s installed costs run above global averages.
• Compare firm cost against firm cost. An intermittent LCOE and a dispatchable LCOE answer different questions. Use IRENA’s firm LCOE concept or model fuel displaced per year.
• Model discipline is procurement discipline. Nine criteria separate a partner who can defend an LCOE model in a lender’s room from one who quotes a number. Ask for the model.

The question of what is LCOE has a simple answer and a difficult application. The formula takes a paragraph. Defending the inputs takes a feasibility study, site data, engineering judgement and honest assumptions about the market you are building in. That is the work USP&E Global does alongside you.

Ready to Power Your Project? Talk to USP&E’s Engineers Free.

USP&E Global offers a complimentary 4-hour engineering consultation for qualified power station, EPC, and O&M projects. Whether you are in early feasibility or ready to mobilize, our team of 350+ engineers across 35+ countries is ready to guide your project to success with speed and without excuses.

USP&E Global. Powering Possibility. Built for the Frontier.

ISO 9001:2015 | ISO 45001:2018 | 150+ Projects | 35+ Countries | Zero Lawsuits

​

Quick Links:

• Contact USP&E by clicking on this link.
• Request a fast equipment quote

Global Resources

• USP&E's Project Experience Portfolio
• USP&E Case Studies and Client Reference Letters
• USP&E Integrity and Compliance policy
• USP&E's ISO Certifications and Credentials