Solar Farm Unit Economics: What Matters Before You Scale

A solar farm is economically attractive when the risk-adjusted, after-tax value of each delivered MWh exceeds the fully loaded cost of producing and delivering it—and when the resulting return justifies the equity capital tied up for several years.

I've built many renewable energy financial planning tools, including one for a solar farm.

For a company planning to build multiple farms, the key unit is not simply “cost per installed MW.” It is:

Cash generated per operating MW, divided by all the capital and development spending required to get that MW into operation—including spending on projects that fail.

The fundamental equations are:

$$\text{Annual net generation} = \text{MWac} \times 8{,}760 \times \text{net capacity factor}$$
$$\text{Operating cash margin per MWh} = \text{realized revenue per MWh} - \text{cash operating cost per MWh}$$
$$\text{Development cost per COD MW} = \frac{\text{development spending on successful and failed projects + platform overhead}} {\text{MW reaching commercial operation}}$$

The most important metrics

Metric	How to measure it	Why it matters
1. Fully loaded capital cost	$/Wdc, $/Wac, and total $/MWac	The main upfront cost. Include development, EPC, owner’s costs, land/site work, interconnection, financing during construction, reserves, and contingency.
2. Interconnection cost and certainty	$/kWac, dollars at risk before certainty, upgrade-cost range	A modestly priced plant with a major network upgrade can be worse than a more expensive plant at a strong grid node.
3. Net energy yield	MWh/MWac-year and  kWh/kWdc-year	Converts installed capacity into saleable production. It should reflect availability, degradation, electrical losses, clipping, curtailment, snow, soiling, and transmission losses.
4. Realized net revenue	$/MWh actually received	More important than the headline PPA or wholesale price. Account for basis, congestion, negative pricing, curtailment, shaping, imbalance charges, RECs, capacity payments, and escalation.
5. Cash operating cost	$/kW-year and $/MWh	Include O&M, land rent, property tax, insurance, asset management, vegetation, security, inverter reserves, grid charges, and decommissioning reserves.
6. Break-even energy price	$/MWh needed to meet target return and debt constraints	This is often more useful than a generic LCOE. It tells you the minimum bankable PPA or merchant capture price the project can tolerate.
7. Return on equity	NPV/MW, unlevered IRR, equity IRR, cash yield, minimum DSCR	Determines whether the farm is genuinely investable after financing, taxes, reserves, and downside cases.
8. Development  productivity	Development cost/COD MW, conversion rate by stage, months to COD	Measures whether your pipeline creates real assets or merely accumulates speculative MW.
9. Equity required per operating MW	Equity invested ÷ retained operating MW	Two projects with the same IRR can have very different capital intensity and therefore very different scalability.
10. Capital recycling velocity	Months from first development dollar to sale, refinancing, or distributable cash	A slightly lower-margin project that returns capital in three years can support more growth than a higher-margin project tying up equity for six years.

1. Fully loaded capital cost

Do not use only the EPC contract price. Build the project cost from the ground up:

$$\begin{array}{rl}{\displaystyle \text{All-in capex}=}& {\displaystyle \text{site control and development}}\\ {\displaystyle }& {\displaystyle +\text{modules, inverters, trackers and BOS}}\\ {\displaystyle }& {\displaystyle +\text{civil and electrical construction}}\\ {\displaystyle }& {\displaystyle +\text{substation, gen-tie and interconnection}}\\ {\displaystyle }& {\displaystyle +\text{owner’s engineering, legal and insurance}}\\ {\displaystyle }& {\displaystyle +\text{construction-period interest and fees}}\\ {\displaystyle }& {\displaystyle +\text{contingency, reserves and initial working capital<span style="text-align: left;"></span>}}\end{array}$$

Always track both DC and AC denominators:

• Modules and many construction benchmarks are quoted in $/Wdc.
• Grid export capacity, revenue, interconnection and project size are often measured in MWac.
• With a 1.30 DC/AC ratio, $1.00/Wdc equals $1.30/Wac before any other adjustments.

Mixing these denominators can make a project appear 20–40% cheaper than it really is.

As a current U.S. reference point, DOE’s modeled market price for a 100-MWdc, single-axis-tracker project in Q1 2025 was approximately $1.066/Wdc in 2024 dollars. Berkeley Lab’s empirical 2024 project data was approximately $1.2/Wdc, or $1.6/Wac. In that dataset, projects larger than 250 MWac averaged materially lower installed cost per MW than projects in the 5–20 MWac range. These figures are benchmarking anchors, not contractor bids or guarantees for a particular site.

2. Interconnection economics

Interconnection deserves its own investment case rather than being buried inside construction cost.

Track:

• Queue and study deposits at risk
• Substation and gen-tie cost
• Network upgrade allocation
• Security postings and letters of credit
• Study restarts and restudies
• Expected energization date
• Probability that the assigned upgrades remain affordable
• Cost of withdrawing from the queue
• Potential curtailment after interconnection

One Berkeley Lab study covering five non-ISO balancing authorities found solar interconnection estimates averaging about $509/kW across all requests and $216/kW among requests with completed studies. The numbers are not national averages, but they illustrate how grid costs can be large enough to overwhelm land or EPC savings.

For early-stage sites, use at least three interconnection cases:

$$\text{Expected interconnection cost} = \sum_s P_s \times \text{Cost}_s$$

For example, a low, base, and severe-upgrade case should each have a probability and schedule implication. Using only the developer’s preferred estimate understates both cost and time risk.

3. Net generation per MW

The best production unit is usually:

$$\text{Net MWh per MWac-year}$$

This captures more value drivers than nameplate capacity alone. It should be based on an hourly production model, not simply annual sunlight.

Your production bridge should show:

1. Gross irradiance and modeled production
2. Temperature and module losses
3. Inverter and electrical losses
4. Tracker availability and mechanical losses
5. Soiling, snow and vegetation effects
6. Plant availability
7. Clipping
8. Grid and economic curtailment
9. Transmission losses to the settlement meter
10. Annual degradation

Maintain both:

• P50: the central expected-production case
• P90: the more conservative case generally used to test debt and downside resilience

A site with higher irradiance can still have worse economics if it faces more congestion, curtailment, extreme heat, difficult construction, or expensive interconnection.

4. Realized revenue per MWh

Do not model revenue as simply:

$$\text{MWh}\times \text{headline market price}$$

Use:

$$\text{Realized revenue per MWh} = \text{energy revenue} +\text{REC and capacity value} +\text{other contracted revenue} -\text{basis, congestion, shaping and imbalance costs}$$

For contracted projects, model:

• PPA price and escalation
• Contracted versus merchant production
• Curtailment compensation
• Availability guarantees
• Production shortfall damages
• Settlement node
• Buyer credit quality
• Change-in-law allocation
• End-of-contract merchant value

For merchant projects, the important metric is the capture price, meaning the average price earned during the hours when the project generates. This can be significantly different from the simple market-wide average price.

A portfolio concentrated in one grid region can suffer correlated pricing risk: all of the farms generate simultaneously, congestion develops, and their capture prices fall together. Geographic and nodal diversification therefore has an economic value even when it slightly increases operating complexity.

5. Operating costs

Track operating expenses in two forms:

$$\mathrm{\$}\mathrm{/}\text{kW-year}\text{and}\mathrm{\$}\mathrm{/}\text{MWh}$$

The first is useful for budgeting; the second is needed to understand contribution margin.

Include at least:

• Scheduled and corrective O&M
• Land lease or land carrying cost
• Property taxes and local payments
• Insurance
• Asset management and reporting
• Vegetation, panel washing and site security
• Communications and monitoring
• Replacement inverter and transformer reserves
• Transmission and market-participation charges
• Decommissioning or restoration reserve

Land should also be translated into energy economics:

$$\text{Land cost per MWh} = \frac{\text{annual lease cost}} {\text{annual net MWh}}$$

Cheap land is not necessarily economical land. A low-cost parcel with weak grid access, lower yield, permitting difficulty or wetland mitigation can be much more expensive per delivered MWh.

6. Break-even price and project returns

LCOE is useful, but it does not fully capture the timing of tax benefits, debt amortization, merchant exposure, curtailment, or contracted escalation.

The more actionable metric is:

What constant or escalated realized energy price makes the project satisfy its target equity return and minimum debt-service coverage?

Calculate at least:

• Unlevered project IRR
• After-tax project NPV
• Equity IRR
• Equity multiple
• Year-one and stabilized cash yield
• Minimum and average debt-service coverage ratio
• Break-even PPA price
• Break-even merchant capture price
• NPV and IRR per MWac
• Downside loss of equity

Run a full sensitivity matrix around:

• Capital cost
• Interconnection upgrades
• Commercial-operation date
• Production
• Curtailment
• Realized price
• Interest rate
• Tax-credit monetization
• Operating costs
• Terminal or merchant value

A project with an excellent base-case IRR but a large equity loss in a modest downside case is usually less scalable than a project with slightly lower upside and much stronger downside protection.

7. The most important portfolio metric: cost per successful MW

When building multiple farms, project-level budgets systematically understate the true economics because they exclude money spent on unsuccessful sites.

Use:

$$\text{True development cost per COD MW} = \frac{ \begin{array}{c} \text{site-control, engineering, queue, permitting and legal spend}\\ +\text{spending on withdrawn projects}\\ +\text{development-team overhead} \end{array} } {\text{MW reaching commercial operation}}$$

For example, suppose you spend:

• $12 million on ten development projects
• Three projects totaling 300 MW reach operation
• Seven projects are abandoned

Your development cost is not $12 million divided by the original pipeline. It is:

$$\$12\text{ million} \div 300\text{ MW} = \$40{,}000/\text{successful MW}$$

That is the number that belongs in the portfolio economics.

Raw pipeline MW should also be risk-adjusted:

$$\text{Risk-adjusted pipeline MW} = \sum_i \text{Project MW}_i \times P_i(\text{reaching COD})$$

Historical queue data demonstrates why this matters. By the end of 2025, only about 13% of the U.S. capacity that entered interconnection queues from 2000 through 2020 had reached commercial operation, while most had withdrawn. Projects completed in 2025 had a median interval of more than five years from interconnection request to commercial operation.

Track conversion rates separately at each stage:

$$\text{Site identified} \rightarrow \text{site controlled} \rightarrow \text{queue submitted} \rightarrow \text{study completed} \rightarrow \text{permitted} \rightarrow \text{offtake secured} \rightarrow \text{financed} \rightarrow \text{NTP} \rightarrow \text{COD}$$

A company with fewer early-stage MW but a much higher queue-to-COD conversion rate can be worth substantially more than a developer with a huge speculative pipeline.

8. Capital recycling

For a multi-project strategy, calculate:

$$\text{Equity invested per COD MW}$$

and:

$$\text{Capital cycle}=\text{date capital is returned}-\text{date first capital is spent}$$

Then calculate annualized portfolio production:

$$\text{COD MW per year per \$1 million of development capital}$$

This distinguishes three business models:

Develop and sell

The core metric is:

$$\text{Expected development margin per MW} = \text{sale proceeds per MW} - \text{risk-adjusted development cost per MW}$$

Time matters heavily. A $100,000/MW margin earned after two years can be more attractive than a $140,000/MW margin earned after five years.

Build, own and operate

The primary metrics are after-tax NPV/MW, equity IRR, annual cash yield, DSCR and long-term realized margin per MWh.

Build, refinance and retain

Measure:

• Original equity per MW
• Cash returned at construction or term refinancing
• Residual equity still invested
• Retained ownership percentage
• Cash-on-cash yield on the residual equity
• Months until capital is available for the next project

This model can compound rapidly when refinancing returns most of the original equity, but it becomes fragile when construction overruns or lower valuations prevent the expected capital release.

A simple illustrative per-MW example

Assume, purely for illustration:

• 1 MWac
• 27% net capacity factor
• $45/MWh realized revenue
• $18/kW-year of total cash operating costs

Annual production would be:

$$1 \times 8{,}760 \times 27\% = 2{,}365\text{ MWh}$$

Annual revenue:

$$\mathrm{2,365}\times \mathrm{\$}45=\mathrm{\$}\mathrm{106,434}$$

Annual cash operating costs:

$$\mathrm{1,000}\text{ kW}\times \mathrm{\$}18=\mathrm{\$}\mathrm{18,000}$$

Operating cash margin before debt, taxes, replacement reserves and corporate overhead:

$$\$106{,}434-\$18{,}000 = \$88{,}434/\text{MW-year}$$

The sensitivities reveal what drives value:

• A $5/MWh change in realized price changes annual cash flow by about $11,826 per MW, or approximately $1.18 million for a 100-MW plant.
• A 5% curtailment loss costs approximately $5,322 per MW-year at $45/MWh, before considering any related congestion or negative-pricing effects.

That is why small changes in capture price, curtailment and production often matter more than modest reductions in land rent.

U.S. tax-credit caution

For U.S. projects, do not automatically place a 30% investment tax credit into every underwriting model. The Section 48E clean-electricity investment credit has a 6% base rate and can generally reach 30% when prevailing-wage and apprenticeship requirements are satisfied, with possible additional bonuses. Eligibility, construction timing, transfer discounts, recapture exposure and monetization timing all affect its actual cash value.

Under the current special wind-and-solar timing rule, an applicable solar facility that begins construction after July 4, 2026 generally cannot claim Sections 45Y or 48E when it is placed in service after December 31, 2027. Projects relying on earlier beginning-of-construction status require careful documentation and continuity analysis. Given that it is now July 2026, this issue should be resolved with tax counsel before treating the credit as part of project value.

The dashboard I would use

For each individual farm, put these at the top of the investment memo:

1. All-in $/Wdc and $/Wac
2. Interconnection $/kWac and upgrade downside
3. P50 and P90 MWh/MWac-year
4. Contract price and realized net $/MWh
5. Cash operating margin $/MWh
6. Break-even PPA or capture price
7. Total equity/MW, equity IRR and minimum DSCR
8. Schedule contingency and downside NPV

For the overall company, use:

1. Risk-adjusted pipeline MW
2. Development spend per COD MW, including failures
3. Stage-by-stage conversion rates
4. Average months from site control to COD
5. Equity capital required per retained MW
6. COD MW per year per employee and per dollar of overhead
7. Capital returned divided by capital invested
8. Portfolio overhead per operating MW
9. Geographic and offtaker concentration
10. Actual-versus-underwritten production, price, cost and schedule

For a first farm, prioritize interconnection certainty, a bankable revenue arrangement and a genuinely buildable site ahead of the lowest apparent land or EPC price. The scalable advantage comes from repeatedly converting development spending into operating MW and recycling capital—not merely from accumulating projects in the queue.

If you have a project you need a custom financial model built for, you can hire me for that.

My financial model template library has over 200 unique industry-specific tools.

Article found in General Industry.