Electrification
Sequencing

Owners often treat electrification and power supply as separate projects. One team replaces fuel-fired equipment; another expands utility service, buys storage, or studies on-site generation. That division is administratively convenient and analytically dangerous. Supply sized against yesterday's load can be wrong on the day it enters service.

Section 01The load decision comes first

Electrification is usually described as a demand problem: more equipment connects to the electrical system, the peak rises, and the owner needs more capacity. That description is incomplete. Electrification changes the conversion technology itself. A fuel-fired boiler, a mechanical drive, a lift truck, a process heater, and an electric alternative do not turn purchased energy into useful work in the same way. Electric demand can rise while total delivered energy falls. Maintenance, local emissions, controllability, product quality, and thermal integration can change at the same time.

The first decision is therefore not how much generation to install. It is what the site will actually do after the operating process changes. The answer has three parts: annual electric energy, the highest credible coincident demand, and the hourly shape between them. California's current demand forecast uses the same basic discipline at system scale. The California Energy Commission separately models energy efficiency, fuel substitution, transportation electrification, known new loads, behind-the-meter resources, and hourly and peak demand rather than collapsing them into one annual total.¹

A site study should be smaller, but not cruder. It should distinguish committed projects from possibilities, flexible loads from fixed loads, and the current process from the future process. Only then can the owner compare utility capacity, load management, storage, solar, firm generation, thermal systems, and the no-project case on a common basis.

Section 02More electricity can mean less total energy

The phrase adding load sounds inefficient because the electric meter moves upward. The meter is not the whole energy system. The relevant comparison is useful output divided by all purchased energy, with the operating boundary held constant.

The National Renewable Energy Laboratory reported those paired results in its 2018 Electrification Futures Study demand-side scenarios. They are national scenario outputs, not a prediction for any California site. Their value here is conceptual: a larger electric numerator can coexist with a smaller total-energy denominator when conversion losses fall.²

The industrial case is more specific. The U.S. Department of Energy's 2025 industrial electrification booklet describes direct electric process-heating technologies including resistance, induction, infrared, microwave, radio-frequency, and electric-arc systems. It also states the limits plainly: suitability depends on process temperature, production rate, heating medium, integration, and location; electrification is not one technology and is not universally applicable.³ That is the right owner-side posture. The question is never whether electricity is better in the abstract. It is whether a defined electric process makes the site's useful output better enough to justify its capacity, tariff, controls, and delivery consequences.

Section 03Build three load cases, not one

A defensible sequence begins with three load cases. The first is the measured present: interval data, production schedule, weather, maintenance outages, and the actual coincidence of major loads. The second is the committed future: projects with approved capital, a real delivery path, and an operating date. The third is the option case: plausible electrification and growth that management has not yet committed.

Mixing these cases creates two opposite errors. If optional loads are treated as certain, the owner can overbuild service and supply. If committed electrification is omitted, the owner can commission a system that is short on capacity immediately. Both errors come from treating nameplates or annual consumption as a load forecast.

Hourly shape matters because two projects with equal annual energy can impose different capacity requirements. One may run steadily. Another may create a short, weather-sensitive, or shift-change peak. Lawrence Berkeley National Laboratory's end-use load-profile work describes hourly profiles as the link between how and when energy is used and the value of efficiency, demand response, distributed generation, and storage.⁴ The California Energy Commission publishes hourly impacts for efficiency and fuel substitution by end use and utility planning area for the same reason.¹

The owner does not need false precision. Each case should identify what is measured, what is vendor-sourced, what is management-committed, and what remains illustrative. A range with honest provenance is better than a single confident number assembled from incompatible assumptions.

Section 04The honest case for electrifying first

Electrification should precede supply expansion when it changes the useful-load basis enough that sizing the old process would waste capital or lock in the wrong operating architecture.

• Process efficiency changes materially. Direct electric heating, variable-speed drives, heat pumps, or electric material handling may reduce conversion losses or idle losses. The project must be evaluated at required output and duty cycle, not by comparing equipment nameplates.
• Control improves. Electric equipment can offer faster modulation, finer zone control, and easier scheduling. That can turn part of a nominally firm load into a flexible one, but only if the production owner accepts the operating envelope.
• Avoided systems have value. Fuel piping, combustion air, exhaust, onsite fuel handling, and some maintenance tasks may shrink or disappear. Their avoided capital and operating burden belongs in the comparison.
• The future process defines resilience. Backup sized for a legacy fuel process may protect the wrong loads. Electrification clarifies which electric functions must ride through an outage, which may pause, and which retain a non-electric fallback.
• Supply can be modularized around a measured ramp. When electrification occurs in phases, service, storage, and generation can follow real operating evidence rather than a single large forecast.

These are reasons to study electrification first, not reasons to approve it automatically. The sequencing advantage disappears if the electric process is technically immature, if the utility date misses the operating deadline, or if a critical thermal load loses a practical fallback.

Do not size the future supply around a process the owner already intends to retire.

Section 05The honest case against electrifying first

Electrification can also be the wrong first move. Some industrial processes require temperatures, heat flux, material interaction, or continuous operation that available electric equipment cannot deliver at acceptable production risk. Some sites have adequate fuel infrastructure but constrained electric service. Some electric tariffs make a new coincident peak more important than the annual energy saved. Some conversions create a controls or training burden that is easy to omit from a vendor comparison.

Distribution capacity is not an abstract concern. The California Public Utilities Commission's Electrification Impact Study Part 2 evaluates primary and secondary distribution upgrades under base, equity-driven, and demand-flexibility scenarios. The proceeding explicitly examines hourly load shapes, infrastructure needs, upgrade costs, and the ability of managed shifting and shedding to mitigate some impacts.⁵ A private owner faces the same logic at the meter: new electric equipment is not ready merely because it can be ordered. The service path, switchgear, protection, operating schedule, and curtailment consequence must work together.

Keeping a fuel-fired process can therefore be rational where the equipment has useful life, fuel supply is firm, heat recovery is valuable, air and safety obligations are manageable, and electric capacity is late or expensive. A hybrid can also be rational: electrify the controllable or low-temperature duty while retaining fuel for high-temperature or resilience-critical work. The correct answer may be delayed electrification, partial electrification, or no conversion. Technology-neutral analysis protects all three possibilities.

Section 06Compare the supply paths after the load is fixed

Once the future load cases are explicit, the supply comparison becomes clearer. Every path has a legitimate case and a boundary.

The comparison should not ask which technology has the lowest nominal energy cost. It should ask which complete path serves each load case by the required date, under credible outage, maintenance, fuel, tariff, and operating conditions.

Section 07Flexibility is designed, not assumed

Electrified load can be an asset to the power plan when its timing can move without damaging production, comfort, safety, or product quality. Vehicle charging may move within a dwell window. Water heating or thermal storage may shift across hours. Batch processes may move between shifts. Some cooling or pumping work can be scheduled. Other loads cannot move at all.

NREL's 2021 operational analysis found that demand-side flexibility, particularly optimized vehicle charging and flexible building and industrial end uses, can shift electricity use and provide operating reserves in modeled high-electrification systems.⁶ That is system-level evidence, not a promise that a particular site has flexible load. The site earns the designation through operating facts: earliest start, latest finish, minimum run time, restart limits, product constraints, thermal inventory, and the consequence of interruption.

A flexible-service tariff, battery dispatch plan, or smaller generator is only as sound as those facts. If an operator must override the control every difficult day, the load was never flexible. If the production team can accept a defined envelope and the controls enforce it, electrification may reduce the amount of firm supply the owner needs to buy. Flexibility belongs in the requirements document and operating procedure, not as an optimistic percentage in a financial model.

Section 08An illustrative sequencing example

Illustrative example only; not a forecast, quote, or site result. Consider a facility planning to replace a fuel-fired process line, add electric material handling, and expand production. Management is also evaluating a utility service upgrade, solar, storage, and firm on-site generation.

If the supply study begins with the current electric meter plus a generic growth factor, it misses the process conversion. If it adds every equipment nameplate, it overstates coincidence. If it assumes the new process is flexible without an operating agreement, it understates firm demand. The disciplined sequence is different:

1. Define useful output.Production rate, temperature, hours, quality limits, restart behavior, and the dated ramp are fixed before energy options are compared.
2. Test the end-use alternatives.The existing fuel process, electric replacements, efficiency measures, and hybrid process are compared at equal output. Vendor figures stay labeled as vendor figures.
3. Build hourly cases.Present, committed future, and optional future loads are separated. Coincidence and weather exposure are modeled explicitly.
4. Document flexibility.Only load that operations can move within a written envelope is treated as shiftable or curtailable.
5. Price complete supply paths.Utility service, controls, solar, storage, firm generation, thermal integration, hybrids, and no-project are tested against the same load cases.
6. Stage the capital.Early equipment and enabling work preserve later choices; later capacity is released when the measured ramp or a buyer-approved commitment triggers it.

In that sequence, “adding load” can beat “adding generation” because the end-use project changes the quantity, timing, and controllability of the electricity that generation would serve. The result may still include generation. It is simply sized after the owner knows what it is for.

Section 09The evidence package an owner should require

A board-ready electrification sequence should be auditable. It needs the following evidence, with an owner and date attached to each item:

• Current interval data, production schedule, fuel consumption, weather context, and outage history.
• A future-load register separating committed, conditional, and optional projects.
• End-use requirements and vendor performance data at the site's actual duty, with assumptions labeled.
• An hourly delta-load profile for each conversion, including startup, standby, seasonal, and worst-day behavior.
• A written utility capacity and timing position, plus the owner-side electrical path from service point to equipment.
• A flexibility register stating what may move, for how long, under whose authority, and with what operating consequence.
• Complete-path comparisons for utility service, efficiency, solar, storage, firm generation, thermal integration, hybrids, and no-project.
• Triggers that release each capital phase, and conditions that would change the recommendation.

The Department of Energy's 2025 Industrial Electrification Assessment Framework follows a comparable logic: inventory existing systems, identify applicable electric alternatives, and evaluate the projects rather than treating electrification as a universal substitution rule.⁷ The owner's study adds the service, tariff, resilience, and supply consequences that sit around the equipment choice.

Section 10Fix the future load before buying the future supply

Electrification is neither automatically a capacity burden nor automatically an efficiency win. It is a redesign of how a site converts energy into useful work. That redesign can raise electric demand, lower total energy, change the peak, create flexibility, remove fuel systems, and redraw the critical-load boundary. It can also expose a service constraint, create a new peak, or introduce process risk.

The owner-side sequence is therefore strict: define output, compare end uses, measure the hourly delta, document flexibility, obtain the utility position, and only then size the supply. Every credible path gets its case for and against. Optional growth stays optional. Vendor figures stay sourced. Illustrative figures stay labeled. The recommendation states what would change it.

Generation may be part of the answer. So may utility service, storage, solar, controls, thermal assets, a hybrid, or no project. The decision becomes defensible when none of those paths is allowed to define the load it hopes to serve.