Power Intelligence

How This Comparison Is Calculated

Five on-site generation technologies are evaluated against the same nine inputs: solid oxide fuel cells, aeroderivative gas turbines, reciprocating gas engines, diesel generators, and solar paired with battery storage. Every technology is scored on identical terms, and the constants behind those terms are published rather than described.

Twenty-year net present value, not undiscounted cash flow. Cash flows are discounted at an 8.0% weighted average cost of capital, adjustable between 0% and 15%. Undiscounted twenty-year sums systematically flatter capital-intensive technologies, because a dollar returned in year 20 is counted the same as a dollar spent in year 0. At 8%, that year-20 dollar is worth about 21 cents.

Equipment lead time is priced, not assumed away. Procurement runs in parallel with permitting; construction follows both. On the secondary market, time to power runs from roughly three months for diesel to seventeen months for an aeroderivative package. Through the new-OEM queue, the same aeroderivative package takes about 39 months. That difference is revenue, and the model counts it.

Availability is calculated on identical terms for every technology. Each option is modeled as an N+1 fleet — the spare unit is purchased, not assumed — and fleet availability is computed from unit count and forced-outage rate rather than quoted from a datasheet. At 25 MW, all five technologies land between 99.81% and 99.96%. Vendor availability claims are not used as inputs.

Fuel cost is derived, not entered. Each technology's fuel cost per kilowatt-hour is its heat rate multiplied by a delivered fuel price: $5.30/MMBtu for natural gas (Henry Hub plus a delivery adder) and $25.48/MMBtu for distillate. Nothing is hand-set.

Full constant tables, sources, and scoring weights are published on the methodology page.

What The Secondary Market Is Worth

$770 million is the twenty-year net present value difference between sourcing a 250 MW aeroderivative fleet on the secondary market and waiting for a new-OEM slot, at a revenue rate of $200 per kW-month. The gap is not equipment price. It is 22 months of production that the new-OEM buyer never gets.

The same comparison at other technologies: roughly $593M for solar plus storage — of which $420M is the federal investment tax credit, which a slower project can miss entirely — $389M for reciprocating engines, and $53M for diesel. Fuel cells show no difference, because no secondary market for them currently exists. At higher revenue rates the gaps widen sharply: at $300 per kW-month, the aeroderivative difference reaches $1.2 billion.

Both sourcing paths are modeled in the tool. Switch the equipment source input and every number moves.

What This Tool Does Not Do

This is a screening analysis, not an engineering study. It does not perform a gas interconnection study, size fuel supply, evaluate a specific site's electrical topology, or determine county-level permitting requirements. Air permitting is modeled at the federal classification level; local district rules, offset markets, and health risk assessment requirements vary and can dominate a project's schedule.

It also does not price a specific transaction. Capital costs are mid-range figures from published sources, not quotes. Actual pricing moves with equipment availability, delivery terms, and negotiation — often by a wide margin on the secondary market.

What it does do is tell you which technologies are worth a real evaluation, what they cost across twenty years on consistent assumptions, and how long each will take to deliver power.

Where To Go Deeper

Each technology in the comparison has a corresponding equipment dossier covering specifications, secondary-market pricing by condition tier, inspection criteria, and known failure points: gas turbines, generators, transformers, switchgear, fuel cells, and batteries. Buyers comparing specific units usually start there after running this tool.

For market context, the data center facility library publishes operator and market records across 26 US markets, including operator-level detail on where capacity is being added. Projects evaluating behind-the-meter generation frequently begin with grid interconnection timelines in their target market, then work backward to on-site options.

FAQ: On-Site Power Comparison

How long does it take to get 25 MW of on-site power?

Between three and seventeen months on the secondary market, depending on technology. Diesel generators are fastest at roughly three months. Reciprocating gas engines run about 8.5 months, fuel cells about 11, solar plus storage about 13.5, and aeroderivative turbines about 17. Through the new-OEM order queue, those timelines extend substantially — an aeroderivative package runs about 39 months. Grid interconnection, by comparison, is project-specific: use the service date your serving utility or RTO has given you.

What does on-site generation cost per kilowatt?

$500 to $10,500 per kW installed, depending entirely on technology. Diesel generators sit at the low end at $500 to $1,000. Reciprocating gas engines run $1,200 to $1,800, aeroderivative turbines $1,428 to $1,606, fuel cells $3,000 to $5,000, and solar plus battery storage $8,700 to $10,500 — the last figure reflecting the oversizing and storage required to serve a continuous load. Capital cost alone is a poor selection criterion: at a 20-year horizon, fuel dominates total cost of ownership for every combustion technology.

Are fuel cells cheaper than gas turbines?

Not on total cost of ownership at current constants. Fuel cells carry a 30% federal investment tax credit under IRC Section 48E and the highest electrical efficiency in the comparison, which narrows the gap, but their installed cost is roughly two and a half times either an aeroderivative package or a reciprocating engine. Where they win is on constraints rather than cost: permit-exempt siting, near-zero criteria pollutant emissions, minimal water use, and a compact footprint. In severe or extreme ozone non-attainment areas, those constraints frequently decide the project.

What is the difference between new-OEM and secondary-market equipment?

Delivery date, primarily. New-OEM equipment enters the manufacturer's production queue, which for large gas turbines currently extends years out. Secondary-market units — surplus, cancelled-project inventory, refurbished, or decommissioned assets — already exist and ship in months. Condition and warranty vary by tier, and refurbished units backed by OEM programs can carry warranty terms comparable to new. The tool prices both paths so the schedule difference appears in the economics rather than in a footnote.

Which technology does the tool recommend most often?

That depends entirely on the inputs, which is the point. Reciprocating engines win most gas-available scenarios on cost, aeroderivative turbines gain at larger capacities, diesel wins on speed for short-duration bridge power, fuel cells win where emissions and permitting bind, and solar plus storage wins where land is abundant and load is not continuous.