What “In the Queue”
Actually Means

Somewhere in California this week, an executive told a board that the company's new site is “in the queue” for electric service, and the board heard a delivery date. What the utility heard was an application. The distance between those two understandings is measured in documents, not months, and this paper maps it one milestone at a time.

Section 01One phrase, two different lines

The power industry operates two queue systems that share a vocabulary and almost nothing else. The first is the generator interconnection queue: projects asking to inject energy into the grid, studied in clusters under federally jurisdictional procedures administered by the grid operator. This is the queue of the famous backlog statistics, and its deposits, study cycles, and withdrawal dynamics were built for power plant developers. The second is the service queue: sites asking to take power as retail customers, processed under state jurisdictional tariff rules by the local utility. Different law, different documents, different clocks.

Nearly every owner reading this paper stands in the second line. A manufacturer expanding a plant, a campus electrifying its buildings, a data hall seeking new service: these are load applications, not generation projects. Yet the folklore migrates freely. Generation queue statistics get quoted at load applicants, load side targets get waved at generation developers, and the word “queue” itself carries an image, borrowed from airports, of a single file that advances one place at a time toward a fixed departure. Almost none of that image survives contact with the actual process.

The load side has now become a policy problem in its own right. In June 2026 the Federal Energy Regulatory Commission opened targeted proceedings to speed the integration of large loads onto the bulk power system,¹ and California has been rebuilding its retail connection machinery since 2023. So the first discipline is simply to know which line you are standing in. Everything that follows is the anatomy of the load line: the lifecycle of one large service application, the names of the documents that mark its milestones, what each commits the utility to, and an honest account of where the time goes.

Section 02The lifecycle, stage by stage

No two utilities use identical labels, and the largest requests travel under different tariff sections than small ones. The anatomy, however, is remarkably constant: application, screening, study, contract, construction, energization. Six stages, each closed by a document, each document committing the utility to something specific and, just as importantly, to nothing more.

Stage one: the application

Everything begins with a service application: the site, the requested capacity, the ramp schedule, the requested in-service date, the load's characteristics, and a site plan. Two things are worth knowing about this document. First, it is the cheapest schedule lever the applicant will ever hold. Complete, internally consistent applications with realistic ramps move; applications with placeholder loads and missing electrical detail burn weeks before any utility constraint is even engaged. Second, its acceptance commits the utility to remarkably little. In California, the Public Utilities Commission's energization decision sets a 10 calendar day average target, and a 45 day maximum, for the utility to approve or deny an application for service.² Approval means the request is complete and enters engineering review. It does not reserve capacity, it does not price anything, and it does not start construction.

Stage two: screening and the method of service

Utility engineering then answers three questions: at what voltage the site should be served, from which circuit or substation, and whether the needed capacity exists today. The output carries names like preliminary engineering assessment or method of service determination. What it commits the utility to is an engineering opinion about how service could be rendered. Its consequential content is a fork in the road: either the request can be served from existing capacity, or it depends on upstream capacity work. Nearly everything about the calendar turns on which branch the request takes, and applicants routinely learn the branch only here, well after filing.

Stage three: studies and detailed engineering

Larger requests then enter paid engineering: study or engineering agreements under which the applicant funds detailed analysis and, later, detailed design. The deliverable is a study report naming the required facilities, an estimate of their cost, and a statement of who bears it. For the largest California loads, requests for retail service at transmission voltage in PG&E's territory now move under interim Electric Rule 30, a framework the Commission approved in July 2025, in which applicants agree to fund necessary transmission infrastructure up front as the condition of a streamlined path.³ A study agreement commits the utility to perform the study, to its stated scope, for the stated fee. It does not commit the utility to build anything the study describes.

Stage four: the extension and facilities contracts

When design is real, cost responsibility becomes contract. In California distribution service the anchor tariffs are Electric Rule 15, covering distribution line extensions from the substation to the secondary transformer, and Electric Rule 16, covering service extensions from that transformer to the meter; those definitions come from the Commission's own energization framework.² The rules allocate costs between utility allowances and applicant payments,⁴ and facilities requested beyond the utility's standard design are handled as special facilities, furnished at the customer's expense under Electric Rule 2.⁵ These agreements commit both parties to a defined scope and a defined split of its cost. They are the first documents in the lifecycle in which the utility's obligation is written against money.

Stage five: the construction agreement

For large loads the pivotal milestone is an executed construction agreement: scope, funding, milestones, and an estimated energization date bound together in one signed instrument. It is no accident that utility investor materials report progress by counting the capacity that has reached executed interconnection construction agreements, a far smaller figure than the demand pipeline described in the same materials.⁶ An executed agreement commits the utility to build the defined scope on the agreed funding. Even here, the date is an estimate. Tariffs and agreements state target dates with standard protections for the utility, and the honest reading is a commitment of scope and effort, not a warranted calendar.

Stage six: construction and energization

What remains is physical: utility construction with its crews, materials, and planned outage windows to cut new facilities into a live system; customer construction on the other side of the property line; local inspections; the meter; and finally energization. The Commission's framework describes the whole sequence in eight steps and is explicit that several of them belong to the customer and to permitting authorities rather than to the utility.² Owners who track only the utility's steps routinely discover, late, that the critical path ran through their own site readiness.

Section 03Where the time actually goes

In September 2024, implementing the Powering Up Californians Act, the Commission set the state's first energization targets for its three large investor owned utilities, and it built them on five years of the utilities' own historical performance data.² Two numbers from that decision bracket the honest range for a load application.

For requests served through standard line and service extensions, the adopted average target is 182 calendar days from the request, with maximum targets of 357 days for distribution line extensions and 335 days for service extensions.² Half a year, as a target, for the ordinary case: that is what the regulator judged reasonable against the historical record. The decision's second table is the one large-load applicants should read twice. Where a request triggers distribution capacity work, the adopted maximum timelines are 684 calendar days for a new or upgraded circuit, 1,021 days for a substation upgrade, and 3,242 days, nearly nine years, where a new substation must be built.²

Read those numbers in both directions. They are targets and bounding cases, not commitments on any specific project, and the Commission has continued to refine the framework in later phases of the proceeding. But they are also an official statement about physics and procurement: when a regulator writes a multi-year outer bound for substation work into a decision whose purpose is acceleration, it is telling you what the underlying reality looks like. The same decision requires the utilities to report on energization performance twice a year and to analyze the constraints behind delays, naming funding, staffing, and equipment availability.² Equipment availability written into a standing regulatory reporting requirement is the institutional acknowledgment that long lead electrical equipment, transformers above all, now sets schedules as often as engineering does.

Three further observations complete the picture. First, those targets govern distribution level work; the largest loads, taking service at transmission voltage, sit under the interim framework in which the applicant's up front funding of transmission infrastructure is the stated price of a faster path.³ Second, a material share of the clock belongs to the applicant. Permits, easements, site readiness, and design stability are customer and permitting authority steps in the Commission's own eight step framework,² and a request cannot move faster than its slowest owner-side dependency. Third, these clocks do not run consecutively by necessity. A well managed application overlaps engineering, contracting, and site work; a passively managed one serializes them.

Section 04Queue position is not a date

A queue position is an ordinal. It records the order in which requests arrived, and order matters at the margin. But converting an ordinal into a calendar requires information the position does not contain: which branch your request took at screening, what upstream work it depends on, which documents have been executed, what money has moved, and what the projects around you do next.

Dates move for reasons that involve no bad faith anywhere. The branch can change: a request screened as servable from existing capacity can emerge from study dependent on a substation after all, and the calendar changes regime with it. The neighborhood changes: capacity plans respond to the portfolio of committed projects, and commitments are made by executed agreements and funded scopes, not by application dates. The applicant changes: a revised ramp, a larger load, or a relocated point of delivery resets engineering that was already complete. And the physical world votes last: crew availability, planned outage windows on a live system, and delivery of long lead equipment all sit between an executed agreement and an energized meter.

A queue position records where you stand, not when you connect. The calendar lives in the documents, and it firms only as scope and money attach.

This is why the utilities' own disclosure practice is instructive. Investor materials distinguish between demand in the pipeline and capacity with executed construction agreements, and the second figure is a small fraction of the first.⁶ The market has learned to price the difference between an application and a contract. Owners should read their own position the same way. The only date worth planning capital around is one written in an executed agreement with funded scope behind it, and even that date is an estimate to be managed, not an entitlement to be assumed.

Section 05The questions that locate you

An owner does not need to become a tariff lawyer to know where a request actually stands. Six questions, all answerable in writing, locate any application in the lifecycle:

• Which document is in force today? Not “we are in the queue,” but the instrument's name: application under review, method of service issued, study agreement executed, extension contract signed, construction agreement executed. The name is the location.
• What is the next document, and what triggers it? Every stage ends in paper. If no one can name the next instrument and its precondition, the request is drifting, not progressing.
• Which branch did screening take? Served from existing capacity, or dependent on upstream work; and if upstream, which category: circuit work, a substation upgrade, or a new substation. The category sets the plausible range of the calendar.²
• What money has moved, and what did it buy? Deposits buy review. Study payments buy analysis. Extension and construction payments buy scope. Matching dollars to documents exposes exactly how much commitment exists on each side.
• Which steps are ours? Permits, easements, site readiness, and design stability sit on the applicant's side of the ledger. The utility's clock cannot run faster than the customer's own critical path.
• What is the estimated date in writing, and on what assumptions? A dated, written estimate with stated assumptions can be managed and re-based as facts change. An undocumented date is a hope wearing a schedule's clothes.

These questions are not adversarial. Utility project teams generally answer all six willingly, because a customer who asks them is a customer whose project will not stall on its own side of the line. The questions fail only when they are never asked, which is the normal case, which is why boards keep hearing “in the queue” as if it were a date.

Section 06When the date fails the deadline

Sometimes the located truth is unwelcome: the documented path lands power after the business needs it. That discovery is the beginning of a decision, not the end of one, and the menu is standard even though the right answer is site specific.

Waiting is a path, and occasionally the correct one. Its cost is the value of everything the load would have produced in the interval, and that cost belongs in writing next to every alternative. Flexible or curtailable service, where a utility offers it, can connect a site sooner in exchange for agreed limits on when and how much it draws; the speed is real, and so is the firmness given up, and the curtailment terms are the economics. Phased energization takes partial capacity now and the balance on a negotiated schedule; it fits loads that genuinely ramp, and its later phases inherit the same upstream risks described in Section 03.

Self supply, as a bridge or as a permanent architecture, is where technology choices enter, and every class deserves its honest sentence. Reciprocating engines are typically the cheapest capacity to buy and the fastest to source, and they are combustion machines whose permits in California's stricter air districts can be slow, conditional, or unavailable, with noise and maintenance outages in the bargain. Combustion turbines and microturbines are compact and reward sites that can use their exhaust heat, and their part load efficiency and combustion permitting argue against them where steady thermal demand is absent. Fuel cells avoid combustion, run quietly, and face materially lighter air permitting in several districts, and they carry higher capital cost, exposure to fuel price, and periodic stack replacement as a real lifecycle line. Linear generators offer fuel flexibility and modular deployment, and as a newer equipment class they shift the diligence burden to fleet operating history and service depth, which a study should verify rather than assume in either direction. Solar with storage burns nothing and buys no fuel, and for continuous loads its capacity factor means it rarely stands alone, with land and roof area binding sooner than most first models expect. Batteries alone can bridge short gaps and firm a limited service, and their duration is finite and their recharge depends on the service that already exists.

None of those sentences names a winner, because none can. The winner is a function of the site's load shape, deadline, fuel access, air district, and capital posture, priced side by side with the documented grid path. That comparison, not any single machine, is the decision product an owner should insist on, from us or from anyone else.

Section 07Hold the place, plan by the documents

The application you already filed has value. It holds order in a process where order matters at the margin, and it generates the written record this paper has been teaching you to read. Keep it alive, keep it accurate, and keep its owner side steps off the critical path. But run the capital plan on the documents, not the position. Before an executed agreement, plan in ranges bounded by the regulator's own tables. After one, manage the estimated date as a project risk like any other, with the same rigor you would apply to a supplier or a lender.

That posture is the reason independent, owner's side study work exists. Our product is the comparison: every credible path to power for one site, the grid path documented from the actual instruments in force, the alternatives priced with sourced and dated inputs, one decision. The utility's process is not an adversary in that work; it is a fact pattern. And the owner who reads it document by document sits at the table the way capacity is actually won: informed, funded, and ready to execute the next instrument the moment it is offered.