+34%
PG&E B-19
Increase in 5 yrs
Increase in 5 yrs
California industrial customers pay the highest electricity prices in the country — everyone knows that. The interesting question is what to do about it.
A long argument, told plainly, for why the cheapest, most reliable place to source 24/7 power for a California industrial site is no longer the grid — and what we’re building behind the meter instead.
For most of the last century, the proposition was simple. A factory, a hospital, a data center, a cold-storage warehouse — any operation that needed continuous electricity — signed an interconnection agreement with the local utility, paid the published tariff, and was rewarded with the cheapest, most reliable kilowatt-hour available anywhere. The utility owned the generation; the customer owned the meter; and the line in between was understood to be the most cost-effective place to draw power that human civilization had yet invented.
That arrangement is no longer true in California. It has not been true for some time, and the trajectory is unambiguous. Industrial tariffs — PG&E B-19, B-20, E-19, SCE TOU-8, SDG&E AL-TOU — have escalated at compound rates that no industrial buyer of any other commodity would tolerate. Demand charges, once a rounding error in the bill, are now the dominant line item for many process loads. Reliability, once a default, now requires a separate budget for diesel generators, batteries, and the operational overhead of riding through public-safety power shutoffs that did not exist as a category fifteen years ago.
The premise of this site is that a 24/7 California industrial operation above roughly 500 kW of continuous electrical load is now systematically better off generating its own power on-site than buying it from the utility — provided the on-site generator is the right one. The remainder of this argument is about what that means, why it is true, and what it costs.
The grid did not fail California industry abstractly. It failed in four concrete, measurable ways — each of which compounds the others.
The first is tariff escalation. Over the last five years, PG&E's flagship industrial tariff has risen by more than a third. This is not weather, not fuel-price volatility, and not a one-time recovery from the pandemic. It is structural — wildfire-mitigation capital recovery, undergrounding, vegetation management, and the cost of integrating a transmission system that was never designed for the load and generation mix it now carries.
The second is the public-safety power shutoff. For an operation that runs continuously, an unannounced multi-day outage is not an inconvenience; it is a process loss, a spoilage event, a missed shipment, a contractual penalty. PSPS days now run into the dozens annually in fire-prone service territories, and the trend line points up.
The third is the demand charge — the per-kilowatt fee assessed on the facility's single highest fifteen-minute peak each month, regardless of how briefly that peak occurred. For process loads with even modest peaks, demand charges now constitute the largest single component of the bill, and they are entirely insensitive to how much energy the customer used in total.
The fourth is the interconnection queue. A new industrial site that wants more capacity, or an existing site that wants to electrify a thermal process, now waits twelve to eighteen months for a service upgrade study — sometimes longer — before any work begins. The grid is no longer expandable on the timeline on which industry plans.
The answer is older than it looks. It is industrial co-generation — reframed for the constraints of 2026 and built on a generator that did not exist commercially when the original co-gen wave crested in the 1980s.
Behind-the-meter generation means the generator sits inside the customer's electrical fence. The kilowatt-hours never touch the utility's distribution wire. The customer pays the on-site generator under a long-term power-purchase agreement at a rate that is fixed, escalated only by a defined index, and structured so that the developer — not the customer — carries the capital cost, the operations cost, and the technology risk.
The economic case for this arrangement is not new. It is the same case that drove the first wave of industrial co-generation forty years ago: a customer with a continuous load can produce its own electricity for less than it can buy that electricity at the meter, and capture the waste heat as a second valuable output, provided the on-site generator is reliable enough to displace the bill.
What is new is that the right generator now exists. The reciprocating engines and gas turbines of the previous co-gen wave were noisy, dirty by modern air-quality standards, maintenance-intensive, and culturally incompatible with the kind of facility — the hospital campus, the food-grade processing plant, the semiconductor fab — that most needs continuous power today. The carbonate fuel cell solves all four of those problems simultaneously, and that is what makes the premise of this site practical rather than merely interesting.
Green energy is only sustainable if it's profitable. The economics of behind-the-meter generation are the only economics that scale.
The carbonate fuel cell platform converts pipeline natural gas — or on-site biogas — to electricity through an electrochemical reaction at the stack. There is no flame. The implications cascade.
Because there is no combustion, criteria pollutants — nitrogen oxides, particulate, carbon monoxide — are not produced at any meaningful concentration. NOx emissions are roughly two orders of magnitude below a permitted reciprocating engine, and below the permitting thresholds of the strictest California air districts. A site that could not host a gas engine on air-quality grounds can host a fuel cell.
Because the generator runs continuously rather than cycling on a daily duty curve, it produces baseload power around the clock, indifferent to weather, indifferent to time of day, and indifferent to the state of the transmission system on the other side of the meter. It rides through public-safety shutoffs without diesel because it was already running before the shutoff began.
Because the electrochemical reaction operates hot, the exhaust gas exits at approximately 700°F — high-grade thermal energy that can be captured for hot water, process steam, or absorption chilling. Combined heat and power on a fuel cell is not a retrofit; it is the native operating mode of the platform. Total efficiency, electrical plus useful thermal, can reach 90%.
And because the platform is fuel-flexible, the same hardware that runs on pipeline natural gas today can run on renewable biogas tomorrow without re-permitting, re-engineering, or re-contracting. A path to a net-negative carbon footprint exists on the same physical asset.
| Parameter | Carbonate fuel cell platform |
|---|---|
| Net electrical output | 1,250 kW per module · modular up to multi-MW |
| Electrical efficiency | 49% LHV (vs. ~33% delivered grid average) |
| Combined efficiency with CHP | Up to ~90% when thermal is captured |
| NOx emissions | 0.01 lb / MWh — below CARB BACT and California air-district limits |
| Acoustic profile | 72 dBA at 10 ft — quieter than a typical rooftop chiller |
| Fuel | Pipeline natural gas, renewable biogas, or on-site digester gas |
| Availability | 24/7 baseload — not weather-dependent — runs through PSPS |
| Footprint | One concrete pad, ~58′ × 42′ per module |
From the first conversation to the first kilowatt-hour, a behind-the-meter fuel cell project follows a predictable sequence. The customer carries no capital and no technology risk; the developer carries both.
Twenty minutes on a call, twelve months of interval data, a one-line diagram of the site's electrical service. We model the load, the bill stack, the available footprint, and the eligible incentives, and we issue a Project Feasibility & Site Assessment. Non-binding, no capex, no broker fee.
Front-end engineering — civil, electrical, mechanical, gas, controls — converges on a fixed scope. The customer signs a long-term power-purchase agreement at a defined $/kWh, with a defined escalator and a defined term. Bcal carries the capital.
The Self-Generation Incentive Program reservation is filed with the host utility against the project's specific equity-resilience or low-carbon track. SGIP capture flows to the project economics, not to a separate consultant.
Local building permit, electrical permit, gas permit, and the air-district authority-to-construct. Because the platform is electrochemical rather than combustion, the air-district pathway is materially shorter than for a gas engine of comparable output.
The module ships factory-built. On-site work is concrete pad, gas tap, electrical tie-in, controls integration, and switchgear. Commissioning is a defined sequence ending in parallel operation with the utility service.
Bcal operates and maintains the asset for the term of the agreement. The customer receives a monthly bill at the contracted $/kWh for the metered output of the on-site generator and continues to receive a residual utility bill only for whatever non-fuel-cell load remains.
The PPA structure is straightforward: the customer pays for delivered kilowatt-hours at a contracted rate. There is no upfront capex, no equipment ownership, no operations responsibility, and no technology risk. The number that matters is the all-in $/kWh delivered, compared honestly against the all-in $/kWh of the utility bill it displaces.
An honest comparison includes everything in the utility bill that the on-site generator displaces: the energy charge, the demand charge, the PCIA exit fee, the franchise fee, the wildfire surcharge, the public-purpose programs charge, and the time-of-use multipliers that load the highest cost onto exactly the hours the customer needs to run. It is the all-in delivered number that matters, not the energy line.
For a 24/7 industrial load on a California industrial tariff, the all-in delivered number from the utility today is in the range of $0.32 to $0.44 per kilowatt-hour, depending on territory and rate schedule, and rising on a five-year compound trajectory of roughly 6% to 9% per year. A behind-the-meter fuel cell PPA is structured at a fixed $/kWh in a defensible band below that number, with a contractually defined escalator that is materially below the utility's observed rate of escalation.
The arithmetic across a twenty-year contract is the part that matters. Even a modest delta per kilowatt-hour, compounded over a decade and a half of continuous operation, becomes a number that materially changes the unit economics of the customer's underlying business. That is the proposition.
| Item | Utility-supplied power | Bcal PPA |
|---|---|---|
| Contract structure | Tariff — revisable by the utility / CPUC | 20-year PPA — fixed price + defined escalator |
| Energy charge | $0.18–$0.26 / kWh, TOU-multiplied | Single blended $/kWh, all hours |
| Demand charge | $22–$34 / kW / month | Eliminated for the displaced load |
| Reliability through PSPS | Customer responsibility (diesel + batteries) | Inherent — generator runs continuously |
| Capital outlay | $0 (pass-through in tariff) | $0 (carried by developer) |
| Escalation | ~6–9% / year observed (5-yr trailing) | Defined contractual index, materially lower |
| All-in delivered cost | $0.32–$0.44 / kWh, rising | Fixed $/kWh, defensibly below |
Twenty minutes on a call. Twelve months of interval data. A one-line diagram of your service. We will model the site, identify the eligible incentive tracks, and tell you honestly whether a behind-the-meter fuel cell is the right answer for your load. If it isn't, we will tell you that, too.
info@bcalenergy.com → Book a 15-min screen call →