The Fuel Question:
Honest Screens for Gas,
Biogas, and Hydrogen Blends

Every path to on-site power that consumes a fuel stands on an assumption about that fuel: that it will be available, at the required rate and pressure, at a quality the equipment can live with, on the date the project needs it. Most generation surprises trace back to that assumption going unexamined.

Section 01The screen that comes before the technology

When a site cannot get grid capacity on its schedule, attention turns to on-site generation, and the evaluation that follows is usually electrical: kilowatts, efficiency, footprint, permits, price. The fuel that feeds the machine is treated as a given. Somewhere behind the building, it is assumed, there is a gas line. Somewhere in the digester, there is biogas. Somewhere in the future, there is hydrogen.

In study work, that assumption is where generation paths quietly fail. The gas service that heats a building was never sized to feed a machine that runs around the clock. The digester that produces impressive gas on paper produces it wet, corrosive, and seasonal in practice. The hydrogen-ready label on a brochure turns out to have no written specification behind it. None of these discoveries is fatal early. All of them are expensive late, after capital is committed and the electrical design is frozen.

A fuel screen is the discipline that prevents this. It is a short, documented set of site facts, established before technology selection: what fuel is available at this site, at what rate and delivery pressure, at what quality, under what service class and contractual terms, by what date, and at what cost to change any of those answers. It is distinct from fuel price risk, a market question treated separately in this series. Availability and quality are physical and contractual questions, and they are answerable today, from the utility record, the meter, and the laboratory.

This paper sets out what that screen documents for the three fuel stories a California owner will actually encounter: utility pipeline gas, on-site biogas, and hydrogen blends. The standard is the one we apply to every technology in this series. Each fuel path gets its honest case, for and against, and no path is credible until its fuel facts are in writing.

Section 02Pipeline gas: capacity at the meter is a documented fact

The most common fuel assumption is the simplest: the site already has gas service, therefore the generation project has fuel. Sometimes that is true. But the existing service was sized for the site's historical thermal load, the boilers, kitchens, and process heat, with diversity factors that assume most equipment is off most of the time. Continuous generation breaks that arithmetic. A machine that converts fuel to power around the clock draws gas at a rate that can dwarf the loads the service was engineered for, every hour of the year.

So the first screen is not a philosophy. It is four documents.

First, the meter facts. The existing meter set's capacity and its delivery pressure, from the utility's records rather than from memory or old drawings. Second, the system behind the meter. Whether the serving main and upstream regulation can support the increment is the utility's engineering determination, not the customer's assumption, and requesting it starts a process with its own clock. Third, the service class. Gas tariffs rank customers for curtailment in a supply emergency; generation load generally takes service in tiers that are subject to interruption before residential core customers, and the specific tier a given service falls in is a documented tariff fact. A generation plan whose fuel can be curtailed should say so on its face. Fourth, the upgrade case. What any new or enlarged service costs, and how long the utility states it will take.

The cost regime behind that fourth document changed materially. Effective July 1, 2023, under Decision 22-09-026 of the California Public Utilities Commission, applications for new gas line extensions no longer receive the allowances, refunds, or discounts that ratepayers historically funded; absent a Commission-approved exemption, the applicant bears the cost.¹ The policy context matters as much as the invoice. California is managing toward lower gas throughput over time, and new gas load is now reviewed as an exception rather than courted as growth. An owner planning a gas-fueled generation path should expect to fund the fuel-side infrastructure, justify it, and treat the utility's written schedule as a first-rank project input.

Delivery pressure divides the equipment classes. Reciprocating engines commonly accept the delivery pressures found on distribution service. Gas turbines and many microturbines require elevated fuel pressure, and where the service cannot supply it, the project adds a fuel-gas booster compressor: capital, a parasitic electrical load, a maintenance schedule, and a new single point of failure, all of which belong in the comparison rather than the appendix. Fuel cells typically accept moderate delivery pressure and are instead the most demanding class on fuel cleanliness, which is the subject of the next section.

The honest case for the pipeline path is strong. Where capacity exists at the meter, fuel arrives without trucks, tanks, or a second construction project, through infrastructure with a century of operating practice behind it. The honest case against is equally plain. New gas load pays its own way, waits on the utility's engineering and construction clock, sits somewhere in a curtailment order, and ties the project to a fuel system whose long-run role in California is a matter of active policy. A screen does not editorialize about any of that. It documents which of those facts apply to this site, and dates them.

The fuel side of a generation project deserves the same scrutiny as the electric interconnection it is meant to bypass.

Section 03What pipeline quality promises, and to whom

Pipeline gas is often described as a commodity. At the meter it is closer to a specified industrial product. In northern California, the quality specifications for gas received into the utility's system appear in the gas tariff itself: Gas Rule 21 requires, among other limits, no more than 1 percent carbon dioxide by volume, no more than 0.1 percent oxygen, no more than 4 parts per million of hydrogen sulfide, no more than 17 parts per million of total sulfur, no more than seven pounds of water vapor per million standard cubic feet at typical delivery pressures, no free liquids at the receipt point, and interchangeability with the gas already in the pipe.²

Two readings of that list matter to an owner. The first is reassurance. Delivered pipeline gas is clean by any combustion standard, and for engines, turbines, and boilers, tariff-quality gas with a standard filtration and knockout skid is a solved problem backed by decades of fleet practice. The second is precision about what the tariff does and does not promise. These are receipt-point specifications: they govern what enters the system. Heating value at a given meter varies by day and by location within interchangeability limits, which is why generation equipment is tuned to a range rather than a number. And the tariff's sulfur allowance, small as it is, includes the odorant that makes leaks detectable. Trace sulfur is a feature of delivered gas, not a defect.

That last fact is where the equipment classes diverge. Combustion equipment burns odorized gas indifferently, but the emission-control catalysts attached to it age with sulfur exposure, and their replacement schedule is an operating cost the model should carry. Fuel cells are stricter by design rather than by degree: the catalysts and electrochemical materials at their core are poisoned by sulfur at concentrations well below tariff limits, which is why fuel-cell installations carry dedicated desulfurization beds as standard equipment. That is not a defect either. It is a designed-for condition, and the screen's job is to price each candidate's treatment train, media replacement included, as ordinary operating cost rather than to discover it in year two.

The rule the screen enforces is simple. Read the delivered-fuel specification and the manufacturer's fuel specification side by side, and price whatever equipment closes the gap between them. For tariff gas into combustion equipment, the gap is small and inexpensive to close. For tariff gas into the most sensitive classes, the gap is known and priceable. For the fuels in the next two sections, the gap is the story.

Section 04On-site biogas: the waste stream sets the budget

On-site biogas is the one fuel story where the molecules are already on the property. At wastewater plants, food processors, dairies, and landfills, anaerobic digestion or decomposition produces a burnable gas that many sites flare today, and turning a disposal obligation into an energy input has genuine appeal: a fuel with no commodity bill, produced at the point of use, from a process the site runs anyway.

The screen's first job is arithmetic. Biogas supply is set by the waste stream, not by the electric load. Digester gas production follows plant loading, feedstock character, and digester temperature; landfill gas follows waste in place, age, and moisture, and declines over the years after a cell stops receiving waste. The federal reference points are stable and public: the U.S. Environmental Protection Agency characterizes landfill gas as roughly half methane and half carbon dioxide with small amounts of other compounds,⁴ and digester biogas as 50 to 75 percent methane, with carbon dioxide, hydrogen sulfide, water vapor, and trace gases making up the balance.⁵ Methane carries the heat, so a gas that is half methane carries roughly half the heating value of pipeline gas, and every piece of fuel-side equipment sizes up accordingly.

Three consequences follow. Generation on biogas is fuel-limited: the machine makes what the gas allows, and if the requirement is continuous electrical output beyond what the gas can sustain, the gap must be closed by blending pipeline gas, accepting the shortfall, or resizing the ambition; each is its own engineering decision. Second, the fuel is variable: production swings with season and process, and a screen built on one optimistic month is not a screen. The evidence standard is at least a year of metered gas production, or a stated, defensible proxy labeled as such, with the trend line, growth or decline, in the document. Third, ownership: the gas has an owner and sometimes an existing contractual obligation, and a generation plan financed against fuel the site does not control in writing is a plan built on someone else's asset.

The screen's second job is chemistry. Raw biogas leaves a digester saturated with water vapor, carrying hydrogen sulfide at concentrations that vary widely by feedstock and routinely sit far above pipeline receipt limits, and, at wastewater plants and landfills, carrying siloxanes: volatile silicon compounds from soaps and personal-care products that survive into the gas. Each attacks equipment in its own way. Hydrogen sulfide forms corrosive acid in condensate and sulfur oxides in exhaust, degrades engine oil, and poisons catalysts.⁵ Siloxanes oxidize in any hot section into hard silica deposits that abrade cylinders, foul turbine blades and recuperators, and blind catalysts and electrochemical surfaces. Moisture condenses in every low point that lacks a drain. None of this is exotic; all of it is priceable once the composition is actually measured. The screen requires a laboratory gas analysis, sampled on more than one occasion because composition drifts, including siloxane speciation, with the cleanup train scoped and priced for each equipment class under consideration.

Class by class, the honest picture: reciprocating engines are the historical workhorse of digester gas, tolerant of medium-Btu fuel and moderate contamination, with the penalty appearing as accelerated wear, tighter oil-analysis cadence, and shorter overhaul intervals, a maintenance cost rather than a mystery. Turbines and microturbines run well on medium-Btu variants and package heat recovery compactly, but they must compress a wet, corrosive gas to elevated pressure, which drives corrosion-resistant metallurgy and serious drying upstream. Fuel cells convert biogas at electrical efficiencies that are typically the highest among the common on-site classes, with a near-zero criteria-pollutant profile that matters in strict air districts, and they stand behind the deepest cleanup train in the field: sulfur and siloxane removal to levels far below combustion tolerance, whose capital and media costs belong in the comparison or the comparison is fiction. Direct thermal use in boilers is the modest path that produces no electricity at all, and where steady thermal demand exists it sometimes wins on simple economics; an honest screen keeps it in the table. Upgrading to pipeline quality for injection under the utility's biomethane interconnection rules is a separate, capital-intensive project with its own specifications and economics. For many sites the real decision is a three-way comparison among on-site power, direct thermal use, and injection, and the screen's job is to keep all three priced, not to assume the answer.

Section 05Hydrogen blends: two questions, not one

Hydrogen enters most fuel conversations today as a single question about whether the equipment is hydrogen-ready. That formulation conflates two questions that have different owners and different clocks, and a screen separates them.

The first question belongs to the utility and its regulator: will the gas delivered at the meter contain hydrogen, at what fraction, and when. California's serious public work on this began with a study required by state law, performed for the California Public Utilities Commission by university researchers and released in 2022. It examined blends of 5, 10, 20, and 50 percent hydrogen in natural gas infrastructure and reached measured conclusions: blends up to about 5 percent appeared generally manageable in the systems studied; higher fractions raised the likelihood of leakage and of embrittlement in steel pipelines; fractions above roughly 20 percent raised permeation concerns in plastic pipe; and real-world demonstration was recommended before any systemwide injection standard.³ Since then, hydrogen blending in California has proceeded as demonstrations and regulatory proceedings rather than as a standard element of delivered gas. For screening purposes the conclusion is plain: verify the current blending status with the serving utility at study time, and treat any future hydrogen fraction in delivered gas as a scenario to be tested, not a plan input to be banked.

The second question belongs to the equipment, and it is answered by class, model, and vintage rather than by principle. Hydrogen burns faster than methane, widens flammability limits, and carries less energy per unit volume, so a blend shifts flame speed, knock margin, and required volumetric flow simultaneously. Reciprocating engines frequently tolerate modest blends within the manufacturer's stated limits, sometimes with tuning or derating; the operative word is stated, because tolerance differs across engine families and control generations. In gas turbines and microturbines, combustor design governs: some current combustion systems accept meaningful hydrogen fractions and others very little; the number that counts is the one in the written fuel specification for the specific model across its operating range. High-temperature fuel cells reform methane internally and can accept some hydrogen within their written fuel specification; systems engineered for pure hydrogen are a different machine class solving a different problem, and a rating for one says nothing about the other. Boilers and burners are often the most forgiving at low fractions, subject to burner tuning and a review of the safety train.

The screen's rule for the phrase hydrogen-ready is therefore procedural. It is a marketing phrase until it appears as a written specification stating the accepted blend percentage across the operating range, the effect on rating and maintenance, and the warranty treatment at that blend. With those three in writing, the claim is real and can be weighed. Without them, it is a brochure.

Both sides of the hydrogen case deserve stating. For: blend tolerance is genuine option value on a long-lived asset in a state whose fuel policy is in motion, and it costs little to demand the written specification today. Against: a full hydrogen fuel supply at a California site today is a fuel-logistics project, delivered or produced on site, with its own availability screen, its own economics, and its own permits. It is a path that can be studied honestly. It is not a checkbox.

The table condenses the fuel lens by equipment class. It is qualitative by design; the numbers belong to specific models and specific sites, and a screen collects them in writing.

Section 06The fuel screen, in seven documents

Assembled, the screen is seven documents. An owner can demand all seven before accepting any generation recommendation, including ours.

1. The meter facts, in writing.Existing service capacity, delivery pressure, and service class from the utility record, including where the load sits in the curtailment order. Not from memory, not from old drawings.
2. The upgrade case, priced by the utility.If any candidate needs more gas than the service supplies: the utility's engineering determination, the applicant-funded cost under current California rules, and a dated schedule.
3. Each candidate's fuel demand, computed.Fuel rate at the site's actual duty cycle, including compression and treatment parasitics, compared line by line against documents one and two.
4. Measured biogas quantity.At least a year of metered production or a labeled proxy, the trend stated, and the gas's ownership and commitment in writing before anything is financed against it.
5. The full composition panel.Sampled more than once: methane, carbon dioxide, hydrogen sulfide, moisture, siloxanes. The cleanup train scoped and priced for every equipment class still in the running.
6. Blend tolerance in the manufacturer's words.For any hydrogen claim: accepted percentage across the operating range, effect on rating and maintenance, warranty treatment. In writing, or it does not exist.
7. The no-fuel paths, priced alongside.Grid service, storage, solar with storage. Fuel-consuming paths must beat fuel-free ones on the same page; otherwise the fuel question has already answered itself.

Section 07A cheap screen against an expensive surprise

A fuel screen is among the least expensive work in a power study. The facts it collects sit in utility records, tariff sheets, meter data, and a laboratory panel, and most of them can be assembled before a single piece of equipment is selected. What the screen buys is proportionate to what fuel surprises cost: services resized after the electrical design froze, cleanup trains discovered after the machine shipped, blend claims found to be brochures when the warranty conversation started.

There is also a structural reason the screen gets skipped. Nobody selling a machine is paid to run it hard. The fuel section is where a generation proposal accumulates caveats, and caveats do not close sales. An independent study has the opposite incentive, which is the point. Bcal Energy is an independent, founder-led California firm; we prepare technology-neutral power readiness studies on the owner's side of the table, and the screen in this paper is the one we run. The fuel question is not a detail of the generation decision. It is the floor the decision stands on.