The Library
Decision Frameworks
21 papersFour Gates Before a California Public-Energy Project Is Ready
An anticipated grant, a 2027 utility tranche, rolling water financing, and state procurement channels are different mechanisms. A source-linked framework for separating program status, applicant authority, technical responsibility, and procurement.
The Cost of Capital in an Energy Decision: Why the Discount Rate Decides the Answer
The discount rate is the owner's standard for time, risk, and competing uses of money. A disciplined method for selecting it, keeping real and nominal cash flows consistent, and testing utility service, efficiency, storage, solar, generation, and the no-project case on equal financial terms.
The Capacity Ladder: Getting More From the Service You Already Have
Billing demand, electrical-equipment ratings, operating peaks, and utility-authorized capacity are different ceilings. A practical sequence for measuring the real constraint, removing avoidable peaks, shaping flexible work, storing what can move, and obtaining the final operating envelope in writing before buying more capacity.
Power Quality: Harmonics, Sags, and the Costs That Never Reach the Energy Model
A typical customer sees around twelve utility voltage sags a year, and a sag is not an interruption — so it appears in no published reliability index and in almost no energy model. The federal reporting form excludes momentary events by rule. National survey data puts the estimated cost per unserved kilowatt-hour of a momentary event at fifteen times the four-hour figure, because the cost is the restart, not the energy. What to measure, who owes which part of it, and the honest power-quality case for and against every path to power.
Reading Third-Party Ownership Offers: What Owners Trade for Off-Balance-Sheet Power
Someone else buys the plant, owns it, runs it, and sells you the output at a rate below the utility’s. The structure is legitimate and often right — but the accounting standard that took effect in 2019 ended what the FASB called one of the largest forms of off-balance sheet accounting, and the tax rules explain why the buyout starts in year six and never at a fixed price. What the host actually trades, in money, site control, and optionality.
When to Re-Run the Study: Revalidating an Energy Decision as Conditions Move
The continent's consensus ten-year demand forecast moved 69 percent in one annual cycle, from 132 GW to 224 GW of summer peak growth. Some study inputs are revised nightly, some monthly, some once a year. Which of them can actually flip a recommendation, the trigger list that says when to revalidate, and the three re-runs worth refusing.
Working With the Utility: Accounts, Escalation, and Getting a Real Answer
California adopted energization targets in 2024, and the reported figures show why owners misread them: an average of 118 utility-controlled calendar days against 307 days end to end on the same projects, both accurate under the published methodology. The two rulebooks that govern adding load and adding generation, the escalation ladder in order, and the remedy the Commission is not permitted to give you.
Thermal Storage: The Overlooked Half of Load Shifting
Federal analysis puts about 74 percent of U.S. electricity in buildings, more than half of it serving thermal demands. Storing the cold instead of the electricity that makes it avoids a second conversion — but one federal study found pre-cooling raised total refrigeration energy about 5 percent on an event day. The honest case for and against, and the tax definition that stops at the building line.
Who Runs the Plant at 2 a.m.: Staffing and O&M for Owner-Operated Power
A quoted maintenance rate is meaningless until you know what it excludes: one published federal source puts the same fuel cell systems at 0.7–2.0 cents per kWh without stack replacement and 3.6–4.5 cents with it. Four staffing models compared honestly, the compliance work that attaches to continuous operation, and who is legally permitted to touch the switchgear.
Redundancy, Priced: N+1, 2N, and What Uptime Actually Costs
The availability percentages attached to redundancy conventions were withdrawn by the body they are attributed to, in 2009. What N+1, N+2, and 2N actually buy, why the redundancy apparatus is itself a leading cause of outages, and how to price a spare against the cost of the interruption it is meant to prevent.
Sizing for a Load That Has Not Arrived Yet
Growth forecasts are the least reliable input in an energy model and the most expensive one to get wrong in either direction. What oversizing now costs under current utility terms, why part-load behavior reorders the technology ranking, and how to price modularity as the option it is.
The Flexible Megawatt: What Load Curtailment Is Worth Before You Build
Before an owner spends capital on capacity, one cheaper question deserves an answer: how much of the load genuinely needs to be firm in every hour. Published research says rarely all of it. What flexibility is worth, what a California tariff actually pays and charges, and where it cannot substitute for power.
The Technology-Neutral Test
Eight paths to power, one honest way to compare them. A decision framework for organizations facing multi-year waits for grid capacity in California, and a seven-question test to run on any energy recommendation before capital moves.
The No-Project Case: Pricing the Decision to Do Nothing
Every path to power is measured against the alternative of doing nothing, yet almost no organization prices that alternative with the rigor it demands of capital requests. How to put the cost of delay, and the honest value of waiting, on paper.
Bridge vs. Base: Sequencing Storage, Generation, and Grid Service
Every path to power has a duration as well as a price. A framework for separating bridge assets from base supply, sequencing the two honestly, and pricing the day when full utility service finally arrives.
The Owner's Side
Free feasibility is never free. Its cost is recovered in the equipment price, and its conclusion is constrained before the first number is run. How fee-for-decision structures purchase honest analysis, and where vendor proposals still earn a place at the table.
Relocation as an Energy Strategy
When moving the load beats powering the site: the full relocation ledger of people, logistics, entitlements, incentives, and time, and a discipline for comparing it against on-site paths without motivated reasoning in either direction.
The Five Facts
Five facts decide whether site-specific energy analysis is worth paying for: a dated deadline, a documented constraint, control of site and meter, authority to sign and pay, and technology-neutral fit. Before they exist, analysis is guesswork with formatting.
Behind the Meter or in Front of It
Self-supply behind the customer meter and export in front of it are two regulatory identities, not two flavors of one project: different interconnection doors, different tariffs, different permits, different economics. Most industrial answers live behind the meter. This paper maps when they should not.
Export or Not: Why Most Industrial Projects Serve Themselves
Every on-site generation project faces one structural fork before any technology is chosen: does it ever push power onto the grid, or never? The answer changes the interconnection, the compensation, the sizing, and the risk. Most industrial projects should answer never, and know why.
The Board Memo: Presenting an Energy Decision to Non-Energy Leadership
The best energy analysis in the company fails at the moment it matters if it cannot survive contact with a board agenda. How to compress a power decision onto one page for leaders who will never read the study, and should not have to.
Market Evidence
6 papersReading the Queue
PG&E's second-quarter 2026 investor presentation reports a 12.7 GW data-center pipeline. Within it, 490 MW holds an executed interconnection construction agreement and 140 MW is in construction. Every number is true. This paper is a method for reading the distance between them.
Powered Land
Vacancy near record lows, preleasing at twice its historical norm, and land trading on megawatts rather than acres. The market evidence that documented electric capacity has become a priced attribute of industrial property, and the documentation discipline that separates real capacity from claimed capacity.
Load Growth vs. Grid Reality: The 2026 Numbers
National forecasters added 92 gigawatts to the ten-year outlook in a single revision. One California utility reports a 12.7 gigawatt data-center pipeline; 490 megawatts of it holds executed interconnection agreements. What the 2026 disclosures say, and what they mean for a single site.
What “In the Queue” Actually Means
An application for new electric service is a position in a process, not a promise of power. This paper walks the lifecycle of a large-load service request, from application to energization, names each milestone document, and explains why queue position is not a date.
Reliability, Measured: Your Circuit's Actual Record
Resilience spending is routinely justified by anecdote: the outage everyone remembers, the storm that made the news. Meanwhile your circuit's actual interruption record is filed, public, and specific. This paper shows where it lives and how to use it.
"Powered" as a Real-Estate Metric: What Time-to-Power Is Worth
Power availability has joined location and entitlements among the variables that price industrial real estate. But "powered" is a word before it is a fact, and the spread between the marketing and the record is where fortunes are currently changing hands. The closing paper of the founding library.
Grid & Tariffs
8 papersFlexible Service: Trading Firmness for Speed
California's utilities now offer to connect large loads years sooner if the load agrees to curtail on command. The date is real, and so is the fine print. What curtailment terms do to project economics, which loads fit, and the questions to ask before signing.
Demand Charges Are Not the Whole Story
The anatomy of a California commercial and industrial electric bill: energy, demand, and fixed charges, time-of-use windows, power factor, and the seasonal clock, with the common misreadings that quietly inflate projected savings before a tariff sheet is ever opened.
Phased Energization: Negotiating Partial Power
Full service on one date is the exception now, not the rule. How phasing conversations with the utility actually work, which terms are negotiable, how to bridge the tranche that is not yet served, and why later phases inherit the queue's risk.
Utility Letters: Getting the Grid Answer in Writing
Verbal estimates, capacity letters, study results, executed agreements: four different instruments, four different levels of commitment. What each one actually promises, how to request the next one up, and why every alternative should be priced against a written grid case.
The Bill After You Build: Standby Rates and Departing-Load Charges
Self-generation does not end the utility bill; it changes what the bill is for. Standby service and departing-load charges are published, knowable, and routinely missing from vendor models. This paper explains what they are, why they exist, and how to model them honestly.
Paralleling With the Grid: California’s Rule 21, Explained for Owners
Any on-site generator that runs while connected to utility wires is operating in parallel, and in California’s investor-owned territories that parallel operation is governed by Electric Rule 21. What the tariff covers, how the screens work, and what non-export designs change.
The Studies Behind a Grid Connection: What Utilities Analyze and Why It Takes Time
Between your application and your energization date sits a stack of engineering studies most owners never see. This paper explains what utility engineers actually analyze, on both the load side and the generation side, and how projects lose months without knowing it.
Sizing Against the Duck Curve: When the Grid Is Cheap and When It Is Not
California's grid is cheapest in the hours solar floods it and dearest in the hours solar leaves it. That shape, the duck curve, should change how behind-the-meter projects are sized. Mostly, it does not, because models still assume a flat world.
Technology Cases
9 papersCombined Heat and Power in 2026: When Waste Heat Changes the Answer
Combined heat and power lives or dies on a question many proposals never answer: where the heat actually goes. A framework for load coincidence, temperature matching, absorption cooling, and seasonal balance, including the honest cases where recovered heat ends up stranded.
Microgrids and Single Assets: Resilience, Honestly Priced
Islanding, controls, switchgear, fuel security, and a permanent testing calendar: resilience is a real product with a real bill of materials. This paper prices it honestly, and shows when a single standby asset serves a site better than a microgrid.
Reciprocating Engines: The Honest Case For and Against
The least fashionable machine in on-site power remains, on several dimensions that decide real projects, the one to beat. A both-ways examination of reciprocating engines: capital cost, delivery speed, service depth, part-load agility, and the California air permit that can decide the answer.
Fuel Cells: The Honest Case For and Against
A machine several California air districts exempt from air permits, running quietly at electrical efficiencies combustion rarely touches. Also the most capital-intensive path to self-generation, exposed to gas prices, with a stack rebuild on the calendar. Both cases, priced honestly, side by side.
Solar and Storage for 24/7 Loads: The Capacity-Factor Problem
A solar array is rated by its best hour; a continuous load is defined by its worst week. Honest, illustrative math on capacity factor, winter output, land intensity, and oversize ratios: where solar-plus-storage is decisive, and where it structurally cannot stand alone.
Linear Generators and New Entrants: Evaluating Young Operating Bases
Newer generation classes arrive with genuine engineering merit and short operating histories. A five-pillar diligence framework for pricing technology-maturity risk honestly: fleet hours, service network depth, parts supply, performance-data provenance, and counterparty durability. Neither hype nor dismissal.
Gas Turbines and Microturbines: The Honest Case For and Against
Compact machines, high-grade heat, and a maintenance model built around scheduled overhauls, set against part-load efficiency losses, hot-day derates, and fuel-gas pressure most sites do not have. Where combustion turbines genuinely earn their place, and where they do not.
The Economics of Waste Heat
Recovered heat is worth the boiler fuel it displaces, and only when a matched load is standing there to take it. Temperature grades, capture fractions, absorption chilling, and the arithmetic that decides when waste heat flips a technology choice and when it rounds to zero.
Hydrogen Blending: What Today's Equipment Actually Accepts
Hydrogen occupies a strange place in energy planning: overpriced in some proposals, dismissed entirely in others. This paper is the sober middle. What blends mean, what equipment classes tolerate today, and what an owner should and should not pay for now.
Fuels & Environment
6 papersAir Permitting for On-Site Generation
Thirty-five local air districts, thirty-five rulebooks. The BACT trigger that decides what control hardware a generator must carry is five times stricter across a county line, and non-combustion equipment often needs no district permit at all. Where your site sits shapes what you should build.
The Fuel Question
Every generation path that consumes fuel stands on an assumption about that fuel. What a disciplined availability screen documents about gas capacity at the meter, biogas quantity and quality, and hydrogen blends, before any technology decision is credible.
Water, Land, and Noise: The Non-Electric Constraints That Kill Projects
Economics gets the meetings, but water, land, noise, fire code, and fuel access issue the vetoes. The site constraints that disqualify technologies before price ever matters, why they surface late, and a screening matrix that catches them in week one.
The Overlooked On-Site Fuel
Thousands of food, beverage, agricultural, and water facilities already make a usable fuel every day, and many flare it. How to measure the resource honestly in quantity, quality, and seasonality, and how the conversion paths compare when nobody is selling a machine.
The Fuel Line: Modeling Gas Price Exposure Honestly
Every fuel-consuming path to power carries a line item no one controls: the price of gas. How to model that exposure with sensitivity bands and dated forward indications instead of a single flattering number, and why electricity-price assumptions deserve the same discipline.
Counting Grid Emissions Honestly
The same generator, on the same grid, can look green under one accounting method and brown under another. Why average, marginal, and time-matched emissions numbers disagree, which question each one answers, and how to read a carbon claim before repeating it.
Study Literacy & Execution
12 papersThe Load Basis
Nameplate ratings describe what a site could draw; annual totals remove the clock. Only interval data shows the shape that sizes the plant. How load shape picks the technology, how California customers pull their own meter data, and where assumed loads fail.
Firm vs. Estimate: How to Read the Numbers in an Energy Study
Every figure in an energy study is one of four things: sourced, quoted, indicative, or illustrative. A tagging discipline for telling them apart, the four places sales models pass estimates off as firm, and a fifteen-minute audit any owner can run.
The Incentive Screen
Every power proposal carries an incentive line, and it is often stale, padded, or both. Current federal law provides 30 percent for qualifying property; everything after the word qualifying is work. Here is the discipline that makes the number defensible.
Reading Vendor Models: Behind-the-Meter Economics Without Guarantees
Vendor pro formas are offers dressed as analyses. Six patterns flatter nearly all of them: availability, escalators, missing standby and departing-load charges, unqualified incentives, blended-rate math, and terminal value. A line-by-line checklist for owners who sign what boards must defend.
Anatomy of a Power Readiness Study
A defensible power study has six parts: measured load basis, complete paths matrix, sensitivity on the assumptions that move the answer, permit and fuel screens, incentive screen at statutory values, decision memo. What each must contain, and what two weeks of bounded work cannot resolve.
The Retrofit Question: Adding Power to a Site That Is Already Running
Tie-in outages, bus ratings, fault duty, rigging paths, and commissioning around production: why integration cost and downtime, not the equipment quote, usually decide whether a brownfield power project pays, and the questions owners should ask before believing any proposal.
EPC Contracting for Owners: Fixed-Price Myths and Real Risk Allocation
Engineer, procure, construct: three words that owners hear as "someone else's problem now." This paper is about what fixed-price actually fixes, where change orders are born, and the three interfaces that decide whether your project lands on budget.
The Years After Commissioning: Pricing Long-Term Service Honestly
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.
Decommissioning and Residual Value: The End-of-Life Line Nobody Models
Most energy models simply stop: year fifteen or twenty arrives and the spreadsheet goes silent, as if the equipment dissolves and the site sweeps itself. What actually happens at end of life, and why silence about it biases the comparison in whichever direction flatters the seller.
The Permit Map: Every Approval an On-Site Power Project Actually Needs
Air permits get the attention. Then a fire-code plan check, a gas service application, or a noise ordinance quietly takes the schedule. The complete approval inventory for on-site power in California, and the sequencing that keeps it from becoming the critical path.
What Insurers and Lenders Ask About On-Site Power
Two third parties discipline every on-site power project whether the owner invites them or not: the property underwriter and, where there is financing, the lender. Both work the same instrument, questions whose answers must be documents, and both reward the owner who assembled the file early.
Measurement and Verification Without Guarantees
Owners are offered two ways to trust an energy project: a vendor's promise about the future, or a measurement plan that will report the truth as it arrives. This paper is about the second, how rigorous measurement and verification works, and what the first actually costs whoever buys it.