Gas Turbines and
Microturbines: The Honest
Case For and Against

Few machines attract less neutral advice than combustion turbines: the parties who sell them describe a compact source of premium heat, the parties who sell against them describe a fuel-hungry compromise, and both descriptions are marketing. The honest case turns on four site facts that most shortlists never examine: what the site does with heat, how flat the load runs, what pressure the gas arrives at, and which air district the machine must live in.

Section 01Two machines, one family

A gas turbine is continuous combustion on a shaft. A rotating compressor pressurizes air into a combustor that burns steadily rather than in pulses; the hot gas expands through a turbine wheel that drives both the compressor and a generator. Because nothing reciprocates, the machine runs smoothly, packs an unusual amount of power into a small and light package, and rejects nearly all of its waste heat in one place: the exhaust stack.

The family spans nearly three orders of magnitude. Stationary gas turbines are available from about 500 kW to more than 300 MW, and industrial facilities have historically applied machines from roughly 500 kW to 40 MW for on-site generation and mechanical drive.¹ Within that range sit two lineages with different temperaments. Industrial, or frame, machines are heavier and less efficient but run longer between overhauls and suit continuous base-load duty. Aeroderivative machines, adapted from flight engines, are lighter and more thermally efficient but cost more and frequently demand higher fuel-gas pressure.¹

Microturbines apply the same thermodynamics at a different scale and with different design choices. Commercial units run from 30 to 330 kW, packaged into integrated modular blocks up to about 1,000 kW.² The turbomachinery is typically a single-stage radial compressor and turbine on one very fast shaft; the high-frequency generator output is rectified and inverted to 60 Hz power, a conversion that costs approximately 5 percent.² Some designs ride on air bearings and carry no lubricating oil at all. The defining component is the recuperator, a heat exchanger that uses turbine exhaust at roughly 1,200 °F to preheat combustion air. It can more than double the machine's electrical efficiency, at the cost of about 10 to 15 percent of power output, and it is the reason a microturbine's final exhaust arrives at 500 to 600 °F rather than 900.²

These are related machines with materially different honest cases. This paper keeps them separate wherever the difference matters.

Section 02The case for: heat grade, footprint, and availability

The strongest argument for a gas turbine behind the meter is not electrical. It is thermal. The representative gas-turbine cogeneration systems in the federal CHP catalog exhaust at 838 to 916 °F, and essentially all of the recoverable heat arrives in that single dry stream.¹ That temperature makes real steam. An unfired heat-recovery boiler on that exhaust produces saturated steam at the pressures industrial hosts actually run, and configured for it, gas-turbine exhaust can raise process steam at conditions as high as roughly 1,200 psig and 900 °F.¹ No other common behind-the-meter prime mover concentrates heat this way. Reciprocating machines split their recoverable heat between exhaust and much lower-grade cooling circuits, and among non-combustion alternatives only the high-temperature electrochemical platforms approach comparable heat quality, with their own trade-offs on capital and lifecycle. For a plant that burns fuel to make steam anyway, the arithmetic changes: the catalog's representative turbine CHP systems reach 66 to 70 percent total efficiency on a higher-heating-value basis when the heat is actually used.¹

The second argument is density. Gas turbines carry one of the highest power-to-weight ratios of any prime mover, and the generating package itself rejects its heat through the stack rather than through condensers and cooling towers, so the power block needs no cooling-water system.¹ On a constrained industrial site, a compact skid and a stack can matter more than a percentage point of efficiency. Microturbines extend the point: quiet, small-footprint modules that can sit in a mechanical yard or on a structure, added in blocks as load grows.

The third argument is combustion quality. Lean-premix combustors are standard across the class. Gas-fired turbines are typically offered at 15 to 25 ppm NOx at 15 percent oxygen without after-treatment, with the best designs near 9 ppm, and selective catalytic reduction takes installations to low single digits where districts require it.¹ Microturbines run cleaner still at the certificate level: commercial units have been certified below 4 to 5 ppmvd NOx against Southern California's standards, among the most stringent in the country.² Within the combustion family, this is the clean end.

The fourth argument is dependability of a particular kind. Recorded availability for gas turbines on clean gaseous fuel runs around 93 to 96 percent, and microturbine fleets, with few moving parts and manufacturer availability targets of 98 to 99 percent, can be configured in multiple modules so that maintenance on one unit never takes the whole plant down.^1,2 Modularity is a genuine operational strategy, not a brochure line: sequence the modules so each runs near its rating, hold one spare, and the fleet degrades gracefully instead of failing whole.

The exhaust stream is the product. A site with no use for the heat is buying the least valuable half of the machine.

Section 03The case against, part one: efficiency away from the rating point

Now the other ledger, and it begins with the number most brochures print largest. Simple electrical efficiency for the catalog's representative gas-turbine CHP systems runs from about 24 percent at the small end of its 3 to 45 MW range to about 36 percent at the large end, higher heating value.¹ Microturbines net 22 to 28 percent HHV after their own parasitics.² Two honest readings follow. First, at behind-the-meter scale, turbines are unremarkable electrical machines: comparably sized reciprocating engines typically post higher simple electrical efficiency, and some non-combustion technologies higher still. A site that values only electricity will rarely find the turbine winning that column, and pretending otherwise is how bad projects get sold. Second, efficiency in this class climbs steeply with size, which means the technology structurally rewards large, steady loads and punishes small ones.

The rating point itself is a best case, and three slopes lead down from it.

Part load. A gas turbine reduces output by lowering its firing temperature, and efficiency falls with it. The catalog's worked example is blunt: at 50 percent output, generation efficiency dropped from 32.8 percent to 24.8 percent on a lower-heating-value basis. Total CHP efficiency held almost unchanged, but only because the energy shifted into recoverable heat, with the power-to-heat ratio falling from 0.70 to 0.46.¹ A host that can absorb that extra heat share keeps the economics; a host that cannot is simply running a worse power plant. Emissions also generally increase at part load, especially at half load and below, which matters to permits as well as to fuel bills.¹ Microturbines soften but do not escape the slope: at 50 percent output, the catalog's example unit gives up roughly 15 percent of its electrical efficiency, declining from about 30 percent to about 25 percent.² The engineering answer is honest and well established: install multiple smaller units and sequence them so each runs near full load. That answer favors the modular end of the class and belongs in the design, not the afterthoughts.

Hot days. Ratings are set at ISO conditions, sea level and 59 °F.¹ As intake air warms and thins, both output and efficiency fall, which means the machine is weakest on precisely the summer afternoons when a California site is likeliest to need it. Inlet-air cooling buys some of the loss back, at capital, water, and parasitic cost.¹ Elevation compounds the derate.² A performance model quoted at ISO and never corrected to the site's design day is not a performance model; it is an advertisement.

The small print between gross and net. Microturbine ratings already absorb an approximately 5 percent power-conversion penalty and the recuperator's 10 to 15 percent power cost; both machines then lose further margin to inlet and exhaust pressure drops, heat-recovery back-pressure, and the fuel-gas compression described next.^1,2 None of these lines is scandalous. Together they are routinely the difference between the proposal's economics and the site's.

One more structural exposure belongs in this ledger even though it attaches to every gas-fired prime mover: the fuel bill. Every point of efficiency deficit is paid for in gas, every year, at whatever the gas price turns out to be. Any serious evaluation carries fuel-price sensitivity as a first-order input rather than a footnote.

Section 04The case against, part two: the pressure the gas arrives at

The most commonly missed disqualifier in this class is not efficiency. It is fuel-gas pressure. A combustion turbine injects fuel into air its own compressor has already pressurized, so the fuel must arrive at the machine above compressor discharge pressure. The catalog's representative systems require 167 to 538 psig at the skid, rising with machine size and pressure ratio.¹ Microturbines, with modest pressure ratios, still typically require 50 to 140 psig.²

Distribution service rarely delivers that. Local feeder lines commonly run 30 to 130 psig, and final distribution lines from 1 to 50 psig.² The gap is closed by a fuel-gas booster compressor, and the booster is not a footnote. It is capital, footprint, noise, and a rotating machine with its own maintenance schedule and outage modes. It is also a parasitic load: from a 55 psig supply, the catalog puts booster power at roughly 51 kW for its smallest representative turbine system and about 1,370 kW for its largest, and even microturbine boosters draw 2 to 50 kW depending on the unit.^1,2 That power comes off net output and net efficiency after the brochure page was printed. And unless the compressor is spared, it stands as a single point of failure in front of the entire plant; sparing it is more capital again.

The discipline here costs almost nothing and is worth more than most of the modeling that follows it: ask the gas utility, in writing and early, what pressure it can deliver at the required flow, on what service, with what firmness, and at what cost to upgrade. One written answer can qualify or disqualify this entire technology class for a site. Discovering the answer after equipment selection is a familiar and entirely avoidable way to lose six months.

Section 05Permitting: combustion equipment with named rules

Turbines are permitted as what they are: combustion equipment. In California, the air district is the operative regulator, and the rules are specific enough to carry names. In the Bay Area, District Regulation 9, Rule 9 governs nitrogen oxides and carbon monoxide from stationary gas turbines.³ In the South Coast basin, District Rule 1134 sets the NOx limits for stationary gas turbines.⁴ A new unit of consequence passes through new source review: best available control technology, emission offsets where thresholds are crossed, source testing, and ongoing monitoring obligations. In the stricter districts, a gas-turbine installation should be planned from day one with selective catalytic reduction and an oxidation catalyst, which bring reagent storage and handling, monitoring systems, added footprint, and a small back-pressure penalty on the machine itself.¹ None of this is exotic. All of it is schedule and money, and it belongs on the project's critical path next to the interconnection and gas-service questions, not discovered behind them.

At the small end, the posture inverts in an instructive way. Districts set permit-exemption thresholds below which a unit needs no district permit, and California pairs that exemption with a certification requirement: electrical generation technologies exempt from district permitting must instead be certified to the state's distributed-generation emission standards under the Air Resources Board's program.⁵ A microturbine below the district threshold is therefore not unregulated; the compliance burden has moved from the site's permit file into the equipment's certificate. For an owner, that is generally good news, because the certification work was done once, by the manufacturer, instead of per-site. It is one of the quiet reasons the smallest members of this family clear regulatory ground faster than their larger siblings.

The honest comparison point: this permitting burden is structural to combustion, and it is a burden the non-combustion alternatives largely do not carry. Weighing it belongs in the technology decision alongside the combustion family's genuine advantages, not after them.

Section 06The maintenance model: scheduled, specialist, priced by the hour

Turbines age by fired hours and by starts, and their maintenance model is correspondingly rhythmic. For gas turbines, routine inspections arrive roughly every 4,000 hours, including borescope examination of the hot gas path, and a major overhaul is needed every 25,000 to 50,000 hours depending on service, restoring the machine to near-original performance.¹ Frame machines run longer between overhauls than aeroderivatives. Frequent starts and stops accelerate both clocks, and extended operation on liquid fuel shortens intervals materially.¹ A turbine bought for flat continuous duty and operated that way earns its maintenance budget; the same machine cycled daily does not.

The commercial form this takes is distinctive. Turbine maintenance is specialist work with a thinner provider network than the engine trades, and owners commonly carry long-term service contracts priced against fired hours, with terms for starts and for response time. Those terms deserve the same negotiation attention as the purchase price, because over a long service life the maintenance stream rivals the capital cost. An overhaul is a planned outage measured in days rather than hours, so the site's power plan during the outage, whether grid supply, a second unit, or scheduled coincidence with a plant turnaround, is part of the original design.

Microturbines rewrite parts of this model in their favor. Design life is estimated at 40,000 to 80,000 hours with overhaul; the parts count is low; air-bearing designs eliminate the lubricating-oil system entirely; and a modular fleet takes maintenance one module at a time while the rest carry load.² The manufacturer availability targets of 98 to 99 percent are targets rather than field statistics and should be treated as such, but the architecture that supports them, few parts and redundant modules, is real.² The counterweight is the service network: the specialist pool is smaller still, and a site far from qualified service should price response time honestly.

Section 07Where turbines and microturbines genuinely fit

Put the two ledgers together and the fit criteria stop being mysterious.

A gas turbine earns its place where a site runs large, flat, near-continuous load and consumes high-grade heat in quantity, most obviously process steam; where space is tight enough that power density matters; where gas service can deliver the required pressure or the project can absorb a properly spared booster; and where the district's control-technology path, usually including SCR, is understood and budgeted from the start. On such a site the turbine is not a compromise. It is the correct machine, and the 66 to 70 percent total-efficiency figures in the federal catalog describe what it does there.¹

A microturbine earns its place at smaller scale: loads served well by modular blocks in the tens to hundreds of kilowatts per unit, thermal demand in the hot-water and low-pressure-steam range that matches 500 to 600 °F exhaust, sites that value redundancy through multiple small modules, distribution gas at workable pressure, and situations where the certificate-based compliance path of a below-threshold unit is worth real schedule.^2,5

Neither machine fits a site whose economics are electricity-only, where higher-efficiency alternatives win the only column that counts; a duty cycle that is peaky, cyclic, or start-heavy; a site with low-pressure gas and no appetite for booster capital and parasitics; or, for larger turbines, a district posture the project cannot afford to satisfy. In those cases the honest recommendation is a different row of the comparison table, and a study that cannot say so was never a study.

On incentives, brevity is the honest position. Under current federal law as of this writing, the investment tax credit for qualifying energy property is 30 percent under 26 U.S.C. §48 and §48E; whether any specific combustion cogeneration configuration qualifies depends on statutory definitions and construction-timing rules that have changed more than once this decade, and the statutory adders that exist must be individually qualified, never assumed.⁶ No credit belongs in a decision model until qualified tax counsel has passed on the specific facts.

Our position on this class is the same as our position on every class. Bcal Energy sells no equipment and carries no margin on any answer, so we have no reason to flatter turbines and no reason to bury them. In study work, the four site facts from the first paragraph decide the question: a documented use for the heat, a load flat enough to keep the machine near its rating, gas at workable pressure or a priced path to it, and a district posture the schedule can carry. When all four hold, this is one of the most settled, well-understood ways to put firm power behind a meter. When any one fails, the honest answer is a different machine, and the owner deserves to hear it early, in writing, from someone with nothing to sell but the answer itself.