The years after commissioning
The purchase decides the sticker price. The next twenty years decide the cost. How each technology class actually behaves over its service life, what long-term service contracts really promise, and the questions to ask before you own the answer.
Commissioning day is the only day a power asset is exactly what the brochure described. Every day after, it is a machine on a maintenance clock, and the clock's price varies more between technologies than the sticker ever did.
Section 01Every technology carries a different clock
Lifecycle service is not a single line item that scales with size. Each class has a distinct maintenance physiology, documented across the federal technology characterizations that remain the most neutral public baseline for operating and maintenance cost structures.1
| Class | The service physiology | The line owners forget |
|---|---|---|
| Reciprocating engines | Frequent, familiar intervals: oil and consumables on running-hour schedules, top-end work at intermediate milestones, major overhauls at long ones. A deep, competitive service market. | Overhaul downtime and its timing: a major rebuild is weeks, and the schedule slot arrives whether or not it suits production. |
| Gas turbines & microturbines | Longer intervals, costlier events: hot-section inspections and rotor life limits dominate; service is more specialized and OEM-centric. | Duty-cycle sensitivity: starts and load swings consume life faster than steady hours; a cycling duty rewrites the interval math. |
| Fuel cells | Low rotating-equipment burden and quiet steady operation, but the defining event is periodic stack replacement, a scheduled, material cost that behaves like partial repowering. | Treating stack life as a warranty afterthought instead of a priced, dated line in the model, and vendor concentration for that service. |
| Batteries | Minimal daily maintenance; the lifecycle events are capacity fade and augmentation, adding or replacing modules to restore rated capability, plus inverter mid-life work. | Cycling and warranty interaction: the operating profile that earns revenue is also the one that consumes warranty throughput. |
| Solar | Lowest service intensity: cleaning, vegetation, monitoring, inverter replacement mid-life. | Performance drift nobody watches: soiling and degradation are silent, and unmonitored arrays quietly underdeliver for years. |
Industry availability statistics collected across large fleets, of the kind reported through generating-unit data programs, confirm the general picture: well-maintained equipment of every class can run reliably, and neglected equipment of every class finds a way not to.2 The technology choice sets the shape of the service burden; the service arrangement sets whether the shape is honored.
Section 02The anatomy of a long-term service agreement
Long-term service agreements, whatever acronym a vendor uses, share an anatomy. Reading one is a matter of finding six organs and checking each:
- Scope. Which components, which labor, which consumables, and, critically, which exclusions. "Full service" agreements differ by entire overhauls in what "full" contains.
- The availability definition. Contracts promise percentages; the definition decides what the percentage means. Which hours count, how planned outages are treated, how force majeure is scoped, and what the remedy is when the number is missed. A generous-sounding percentage with a narrow definition and a small credit is a decoration, not a commitment.
- Escalation. Fixed escalators compound quietly: a routine-sounding annual percentage takes the fee far above its starting point across a twenty-year term. Model the whole curve, not year one.
- Major-event treatment. Are overhauls, stack replacements, or augmentation inside the fee, priced as scheduled extras, or left to future quotation? The third option converts your biggest lifecycle costs into future negotiations held without leverage.
- End-of-term condition. What state must the equipment be in at handback or renewal, and who proves it? Silent contracts here donate the last years' deferred maintenance to the owner.
- Exit and transfer. Termination rights, assignment on sale of the site, and what happens if the service provider itself falters. For young technology classes, provider durability is part of the technology risk.
An availability percentage is worth exactly its definition, and a service fee is worth exactly its exclusions.
Section 03Comparing technologies on lifecycle terms
Because the service physiologies differ, comparing technologies on capital cost alone is structurally wrong, and comparing them on vendor-quoted service fees alone is barely better, since scopes differ. The neutral method is a levelized service cost: for each candidate, lay every scheduled event on a twenty-year timeline, price the events from the written service offer plus quoted extras, add a labeled allowance for the unscheduled, and divide by expected output. Label the result illustrative, because it is; then let the candidates compete on that number alongside capital, fuel, and permitting.
Done honestly, this exercise redistributes advantage in unfashionable directions. The cheap machine with expensive decades and the expensive machine with quiet ones trade places more often than either's sales material suggests. It also surfaces the operational question owners skip: who on your staff, or under what contract, actually executes the routine layer, because a service agreement that assumes owner-performed dailies is priced for a diligence your team may not have signed up to provide. Public installation databases exist partly because operating reality across thousands of sites diverges from spec-sheet assumptions in exactly this layer.3
The comparison, worked once with illustrative numbers
Two candidates, every figure illustrative. Machine A costs three million dollars installed and carries a service fee starting at eighty thousand dollars a year, but its physiology schedules a six-hundred-thousand-dollar overhaul in years eight and sixteen, outside the fee. Machine B costs four million installed with a fee starting at one hundred forty thousand, majors included. On sticker price and year-one service, A wins twice. Now run twenty years. A's fees at a three percent escalator total roughly two point one million, plus one point two million of overhauls: about three point three million of service life, on top of its three million entry. B's fees at the same escalator total roughly three point eight million, with nothing further: three point eight on top of four. Twenty-year totals: six point three million against seven point eight, and A still wins, but the gap has narrowed from "obviously" to "run the sensitivities." Change one assumption, a five percent escalator on A's uncapped overhaul pricing, or one unbudgeted major, and the ranking flips. That flip, invisible at the sticker and decisive over the life, is the entire argument for levelized comparison.
The escalator deserves its own respect in this arithmetic, because compounding hides in plain sight: three percent a year multiplies a fee by about one point eight over twenty years; five percent multiplies it by more than two and a half. A negotiation that trades one point of escalator for a slightly higher starting fee is often the owner's best trade in the entire agreement, and it is routinely left on the table because year one is the only year anyone models. Whoever writes your comparison should be able to show you the fee curve, all twenty points of it, for every candidate, with the majors marked on the timeline where they land. If that chart does not exist, the comparison has not happened yet.
Section 04The reserve discipline
Whether or not a service contract wraps the major events, the owner's ledger should. The discipline is a maintenance reserve: an annual accrual sized to the scheduled majors, overhauls, stacks, augmentation, inverter replacement, so the costs arrive as planned withdrawals rather than emergencies. Lenders and sophisticated boards ask for exactly this, and its absence in a vendor pro forma is one of the quiet flatteries this series keeps finding: a model with no reserve line is claiming, silently, that year twelve never comes.
Section 05Seven questions before signing the service years
- What, by name, is excluded from the fee, and what did those items cost the last three customers who needed them?
- What is the availability definition, hour by hour, and what is the actual remedy when it is missed?
- What does the fee total across the full term at the stated escalator, and what does the levelized service cost per unit of output come to, labeled illustrative?
- Are the major events inside the fee, scheduled extras with prices, or future quotations?
- What condition must the asset be in at end of term, and who measures it?
- What happens to the agreement if the provider is acquired, exits the market, or fails, and what parts and documentation escrow exists?
- What does the agreement assume my own staff performs, and have they agreed?
The service years are where technology marketing goes to be audited. Price them before you choose the machine, and the machine you choose will owe you fewer surprises.
Sources
- U.S. Environmental Protection Agency, Catalog of CHP Technologies (technology characterization sections, including operating and maintenance cost structures). epa.gov. Accessed August 10, 2026.
- North American Electric Reliability Corporation, Generating Availability Data System, Data Reporting Instructions (fleet availability statistics framework). nerc.com. Accessed August 10, 2026.
- U.S. Department of Energy, Combined Heat and Power Installation Database (operating installation records). energy.gov. Accessed August 10, 2026.
One paper. Every day.
The Bcal Energy White Paper Series covers the decisions, technologies, and market evidence behind time-to-power. New research publishes continuously in the library.
Browse all papersPrice the decades, not the sticker.
The Power Readiness Study is our fixed-fee written analysis of every credible path to power for one specific site: $25,000, technology-neutral by design, sold with no equipment margin behind it. A free 20-minute conversation comes first.
info@bcalenergy.comAbout Bcal Energy. Bcal Energy is an independent, founder-led California firm. We prepare technology-neutral power readiness studies for organizations facing time-to-power decisions, on the owner's side of the table. We sell the decision, not equipment. Author: Bharath Ramanidharan, Founder. Contact: info@bcalenergy.com.
Disclaimer. This paper is general information, not engineering, legal, tax, or investment advice, and not an offer of services on any specific terms. Figures described as illustrative are estimates. Statutory, tariff, and program references are current as of the publication date only; confirm status with qualified counsel and advisors before acting. Bcal Energy provides no guarantee of savings, output, performance, or timelines. © 2026 Bcal Energy.