Replacing a Federal Pacific Stab-Lok panel

How to tell FPE Stab-Lok from the doorway

Start here, because a quote to “replace the Federal Pacific” is wasted money if the panel is something else. Home inspectors, listing agents and contractors routinely use “Federal Pacific” as a synonym for any grey 1970s load centre. It is not. You can do the identification with the hinged door open and the dead-front still on. Photograph the door, the inside of the door, and a straight-on shot of the breaker faces. That is enough for an electrician to form a working hypothesis. It is not a diagnosis, and it is not permission to unscrew the inner cover.

The door, closed. A grey steel enclosure, typically in a garage, basement, utility closet or on an exterior wall of a house built or last rewired from the 1950s through the early 1980s. Some covers are embossed “Federal Pacific Electric” or “FPE.” Many are not. Absence of an outside stamp is not a clearance. Presence of an outside stamp is a reason to open the door, not a reason to skip the label.

The door, open — look at the inside of the door, not behind the breakers. This is where the manufacturer’s label almost always lives. Read it. The words that identify this product line are Federal Pacific Electric, FPE, Stab-Lok, Federal Electric, or, on Canadian equipment, Federal Pioneer. Private-label cousins exist — EMI, American Circuit Breaker (ACBC), and some Challenger-branded Stab-Lok breakers — but the Stab-Lok name on a breaker face is the giveaway even when the door label has peeled. If the label is gone and the breakers do not say Stab-Lok, stop. A photograph to a licensed electrician is the next step, not a screwdriver.

The breaker faces, dead-front still on. Two columns of narrow, blocky moulded-case breakers. Original Stab-Lok toggles are often thinner than a modern Square D or Eaton handle. Many carry a red or orange stripe across the handle, or red-painted tips; not all do, so colour alone is not a test. The word “Stab-Lok” is frequently printed on the breaker itself, or on a label near the top between the two rows. Type markings you may see without unscrewing anything: NA (full-width plug-in), NC (half-width), NB (bolted to the bus). Aronstein’s 2024 report treats NA, NC and NB as the same internal mechanism. Cincinnati Insurance’s 10 July 2025 loss-control note uses the Stab-Lok word, red-painted handles, and a white numbering strip as field identifiers. Treat them as clues, not a verdict.

The directory strip. A white numbering band between the two rows, often with rounded corners and tick marks between circuit numbers, is a common FPE layout cue. Non-Stab-Lok FPE bolt-on equipment, Cincinnati says, is not painted red but still has the same physical shape, and its numbering scheme is white lettering on a black background. That bolt-on line is still FPE. It is not a clearance.

No single main breaker. Some FPE load centres are split-bus, “rule of six” panels: up to six double-pole breakers at the top feed the rest of the panel, and there is no main. You can see that with the door open. It is a design, not proof of FPE — other brands used split-bus too — but on an FPE it matters, because a jammed branch breaker then has no main behind it. The 2020 National Electrical Code narrowed the six-disconnect rule and added an outdoor emergency-disconnect requirement at 230.85. Whether that applies to this replacement is the edition your jurisdiction adopted, not the latest book on a shelf.

What you cannot tell from the doorway. Bus-bar stamping (E-bus or F-bus cutouts), corrosion on the bus, heat damage behind a breaker, double-taps, and whether the grounding electrode conductor is even landed, all require the dead-front off. That is electrician work. This page will not publish steps for removing the cover.

What the CPSC actually did — and what a petition is not

Contractor copy often says “CPSC recalled FPE” or, the opposite error, “CPSC cleared them.” Neither is what the Commission published. A third error is to call the 1980 opening a consumer petition. It was not. The Consumer Product Safety Act lets the Commission open a file from a manufacturer’s report under section 15(b), from its own initiative, or from a petition asking it to issue, amend or revoke a rule. The FPE residential file opened the first way.

CPSC press release 83-008, originally 3 March 1983 and still the Commission’s public statement on this product, says so in the second paragraph: “The Commission investigation into Federal Pacific Electric (FPE) circuit breakers began in June 1980, when Reliance Electric Co., a subsidiary of Exxon Corporation and the parent to FPE, reported to the Commission that many FPE circuit breakers did not fully comply with Underwriters Laboratories, Inc. (UL) requirements.” That is a manufacturer report of non-compliance, not a consumer petition under 16 CFR part 1051. We did not find a Federal Register notice granting a petition to recall residential Stab-Lok equipment. Docket 81-7 — “In the Investigation of Federal Pacific Electric Company and Reliance Electric Company,” order dated 19 October 1981 — is the Commission’s own caption for that investigation. It confirms the file existed as a Commission matter. It is not a granted petition, and it is not the 1983 closing statement.

The investigation, the 1983 notice says, “focused primarily on 2 pole residential circuit breakers manufactured before Reliance acquired FPE in 1979.” Commission testing “confirmed that these breakers fail certain UL calibration test requirements.” UL 489, the standard under which a residential moulded-case breaker is listed, requires the breaker to hold its rated current and to trip within specified times at overloads, including 135% of rating. A breaker that will not do that is, in the language of the standard, defective. The Commission said it was “concerned about the failure of these FPE breakers to meet UL calibration requirements” and, in the next clause, “unable at this time to link these failures to the development of a hazardous situation.”

Reliance’s position, as the same notice records it, was that the calibration failures “do not create a hazard in the household environment”; that FPE breakers “will trip reliably at most overload levels unless the breakers have been operated in a repetitive, abusive manner that should not occur during residential use”; and that at the few overload levels where they might fail to trip under realistic use, “currents will be too low to generate hazardous temperatures in household wiring.” CPSC staff, the notice says, had “insufficient data to accept or refute Reliance’s position.” A full assessment, staff estimated, would cost several million dollars against a $34 million fiscal-year 1983 budget. The Commission closed the file, reserved the right to reopen, and said it would continue investigating circuit breakers generally.

The consumer advice in that 1983 notice is generic. Know which outlets sit on which circuit. Do not overload. Comply with local building codes. Disconnect a product that is failing. Investigate why a breaker tripped rather than just resetting it. That list is written for “all circuit breakers and fuses.” It is not an FPE-specific replacement order, and it is not a finding that FPE equipment is safe to leave in place.

On 18 February 2011, CPSC staff added a boxed note at the top of the same press release. Electricians, homeowners, inspectors and real-estate agents should read it carefully. “The press release announces that the Commission closed the matter without making a determination as to the safety of FPE circuit breakers or the accuracy of the manufacturer’s position on the matter.” That is the Commission’s own gloss, twenty-eight years later. It is not a finding of safety. It is not a recall. It is not the grant of a petition.

People have asked the Commission to reopen. Jesse Aronstein wrote to Richard Stern in CPSC’s Office of Compliance on 7 March 2006, arguing that failing FPE breakers were contributing to fire and injury losses and that the 1983 release was being misread as a clean bill of health. He submitted further test results around 2010. The 2011 staff note is the public residue of that pressure. In September 2019, Representative Paul Tonko and Representative Jan Schakowsky wrote to CPSC leadership asking the agency to review new evidence and to consider a consumer recommendation or a recall. A congressional letter is not a petition grant. CPSC has not published a later closure that finds the breakers safe, and it has not published a recall of residential Stab-Lok. “Not recalled” and “found safe” remain different sentences.

HyreElectrical analysis: the legal posture of a residential Stab-Lok panel in 2026 is this. No US federal rule requires you to remove one that is already installed. CPSC never issued a recall. The 1983 closure is not a safety finding, and the 2011 note says so in terms. Independent testing against UL 489, a New Jersey consumer-fraud judgment about how the listing was obtained, and named-carrier loss-control advice are the reasons electricians and insurers treat the equipment as a replacement conversation. Those are different authorities. They should not be mashed into “the government recalled them.”

What the tests showed — and what they did not

UL 489, the standard under which a residential moulded-case breaker is listed, requires the breaker to trip at or below 135% of rated current within a stated time. A breaker that will not do that is, in the language of the standard, defective. CPSC’s 1983 notice states that Commission testing confirmed FPE breakers fail certain of those calibration tests. The press release does not publish the percentages. The percentages below are from the CPSC-contracted Wright-Malta report as compiled in Jesse Aronstein’s 29 March 2024 technical report, from CPSC’s own laboratory sample as he tabulated it, and from Aronstein and David Carrier’s later field-sample testing.

Wright-Malta tested 122 FPE two-pole breakers, 30 A to 80 A, at 135% and 200% of rated current, on individual poles and on both poles, before and after 500 handle operations — a shortened version of the UL mechanical-endurance test. Ratings in that sample: 30 A (30 breakers), 40 A (35), 50 A (20), 60 A (7) and 80 A (30). At 135% on individual poles, 51% failed to trip before wear and 65% after 500 operations; with both poles loaded, the figures were 25% and 36%. At 200% on individual poles the failures were 1% before wear and 10% after; with both poles loaded, 0% before and 1% after. The safety finding that matters most in that report is jamming: some two-pole Stab-Lok breakers would not open at all on a single-pole overcurrent, because of friction in the common-trip mechanism. A jammed breaker with contacts closed is, as Aronstein puts it, the analogue of a penny behind a fuse.

FPE two-pole breakers below 90 A are, in that report’s description, essentially two single-pole breakers ganged with linked handles; most have an internal common-trip mechanism intended to open both poles if one trips. Older Stab-Lok two-pole units do not have that mechanism. FPE claimed, CPSC recorded, that jamming was a consequence of the test conditions — the toggle operations — and would not occur in actual use. Field samples taken out of houses later disproved that. The friction change that jams the mechanism occurs under ordinary ageing in residential installations. Wright-Malta’s after-500-operations results also went the wrong way for the “exercising the toggles keeps them from sticking” advice: failure rates rose after mechanical cycling, not fell.

CPSC’s own laboratory sample, as tabulated by Aronstein, was small: 14 single-pole breakers (4 failed at 135%; 1 critical failure) and 27 double-pole (20 failed at 135%; 5 critical). He also reports that 85% of those double-pole samples and 39% of the single-pole samples failed one or more UL test criteria. The double-pole units that failed to trip at 200% of rated current were treated as critical safety failures, based on CPSC-sponsored analysis at the National Bureau of Standards (now NIST) of fire-ignition potential in residential wiring. Treat those percentages as a small-sample CPSC lab result, not as a census of every breaker in a house.

Later independent testing of breakers removed from homes is larger. Aronstein’s 2024 compilation (results as of 27 October 2017) reports 1,726 FPE single-pole breakers with a 13% failure-to-trip rate at 135% of rated current, and 718 double-pole breakers with a 42% rate; 8% of the double-pole sample jammed. Combination breaker/GFI units in that table are a tiny sample (eight units) with a very high failure rate — too small to quote as a product-line statistic, large enough to refuse any claim that “only the two-pole breakers were the problem.” The 2024 preface puts the all-brand Stab-Lok-type sample above 3,000 breakers.

FPE’s own test data, and the Southwest Research Incorporated work done under contract to FPE/Challenger, were never made public. Aronstein records that there is no indication those hidden results differ in kind from the CPSC and Wright-Malta numbers. UL itself did not independently test production FPE breakers for listing or follow-up; UL inspectors monitored testing done by FPE personnel at FPE’s facilities. UL, Aronstein writes, claimed to be unaware of the fraudulent testing. The listing mark on a Stab-Lok breaker is therefore not evidence that a third party verified the calibration. It is evidence that FPE was authorised to apply the mark.

A pink UL label or a white dot on the handle is not a clearance either. Aronstein tested 428 of those later-production breakers; 12% still failed the 135% requirement. There is no year or paint mark that clears the line.

HyreElectrical analysis: none of this is a CPSC finding that a given house will catch fire. CPSC said, in terms, that it could not link the calibration failures to a hazardous situation with the data it had in 1983, and that it was not going to spend the money to get that data. The literature since then is independent engineering testing against the same UL 489 trip requirement, plus a New Jersey consumer-fraud judgment about how the listing was obtained, plus a modelled fire-loss paper. Those are different claims. They should not be mashed into “the government recalled them,” and they should not be mashed into “2,800 fires a year, says CPSC.” CPSC has never published a fire, death or dollar attribution for FPE Stab-Lok.

Three failure modes the tests actually name

A Stab-Lok conversation that stops at “they don’t trip” is incomplete. The published engineering names at least three distinct problems, and they are not interchangeable.

Failure to trip at 135% of rating — the overfuse analogue. UL 489’s 135% calibration is the line between a breaker that is doing its job and a breaker that is, in effect, a larger fuse than the circuit was designed for. Aronstein’s example: a 30 A breaker that actually trips at 44 A is functioning the way a 40 A breaker would. Installed on 10 AWG copper, that circuit is essentially overfused. Overfusing is universally treated as unsafe. It is not as dangerous as a breaker that will not open at all, and it is not a CPSC finding that the house is on fire. It is a documented, high-rate departure from the listing standard.

Jamming with contacts closed — the penny-behind-the-fuse analogue. On two-pole Stab-Lok breakers, friction in the common-trip mechanism can prevent the contacts from opening on a single-pole overcurrent, no matter how high the current goes. Wright-Malta identified that mechanism in 1982. Field samples confirmed it occurs in houses, not only after laboratory handle-cycling. Aronstein’s 2017 compilation puts the jam rate at 8% of 718 double-pole field samples. A jammed two-pole breaker feeding a dryer, a range, or a heat pump is a circuit with no overcurrent protection. Split-bus FPE panels, which have no main behind those branch breakers, make that jam the last device on the path.

Bus-bar interconnections inside the panel. CPSC contract CPSC-C-81-1455, Wright-Malta, 20 May 1982, is a failure analysis of a residential FPE panel that ignited at bus-bar interconnections on the back of the panel — not at a branch breaker. Aronstein’s 2024 report treats deteriorating current-carrying joints inside common Stab-Lok load centres as a separate, panel-level hazard: overheating when current flows, and in the worst case fire inside the enclosure. That is why clipping new breakers onto the old bus does not address the whole finding. The bus is part of the product.

Type NB “bolted” Stab-Lok breakers are not a workaround for the stab connection. They bolt to the bus instead of clipping. Aronstein’s 2024 revision is explicit that NA, NC and NB share the same internal mechanism and the same substandard trip performance, and that NB should be replaced with the rest. Cincinnati Insurance, independently, recommends treating FPE bolt-on equipment the same as Stab-Lok for replacement, because the failure-to-trip exposure remains even if the stab connection does not.

The New Jersey case was about testing and labelling, not a national safety ruling

In Yacout v. Federal Pacific Electric Co., Superior Court of New Jersey, Law Division, Middlesex County, Docket No. L-2904-97, the court determined that FPE had violated the New Jersey Consumer Fraud Act. Partial summary judgment is dated 15 August 2002, Judge Bryan D. Garruto, J.S.C. The class-action notice states the violation in one sentence: FPE “cheated during its testing of circuit breakers in order to obtain Underwriters Laboratories (UL) approval,” and “knowingly and purposefully distributed circuit breakers which were not tested to meet UL standards as indicated on their label.” The class was New Jersey persons and entities who owned, used or possessed FPE breakers sold or manufactured from 1965 to 1980 bearing a UL listing mark, plus those who had replaced such breakers since 1991.

That is a civil finding about testing and marketing conduct in one state. It is not a criminal conviction, not a CPSC recall, and not a ruling that every Stab-Lok breaker in the country is unsafe. Settlement relief was limited to that New Jersey class; published accounts of the 2005 settlement put homeowner relief at $500 per panel, with institutional plaintiffs receiving varying amounts. There is no remaining manufacturer warranty programme in the United States for residential Stab-Lok equipment. A New Jersey class notice from twenty years ago is not a cheque you can cash in 2026, and it is not a federal order to replace the panel.

Hanover Insurance’s risk-solutions note (Hartford Steam Boiler material, LC February 2019) cites that 2002 partial summary judgment by docket number, then draws its own loss-control conclusion: because of the fraudulent UL testing, the original Stab-Lok panelboards and breakers “were never verified as suitable for the intended use,” and “the safest course of action is to replace it with a completely new panelboard and breaker installation.” That is a named carrier’s advisory position, citing the New Jersey finding. It is still not a CPSC recall.

Aftermarket Stab-Lok breakers, Federal Pioneer, and the other names on the bus

Once FPE left the residential breaker business in the mid-1980s, other companies made breakers that clip into a Stab-Lok bus. Challenger, American Circuit Breaker, and later UBI / Connecticut Electric have all sold “replacement” Stab-Lok-type breakers. The marketing claim is that you can keep the old can and refresh the breakers. The testing claim, on the samples we have, is the opposite.

Aronstein’s 2024 compilation found UBI Stab-Lok-type breakers failing the 135% trip test at higher rates than the FPE breakers they are sold to replace, including brand-new units purchased at retail — 194 of 466 (42%) failed at 135%, and 17 (4%) jammed. The 2024 preface is blunt: replacing FPE breakers with other brands of Stab-Lok-type breakers “may actually increase the risk of an electrical fire,” and “the safety problem posed by the defective Stab-Lok breakers cannot be alleviated by replacement breakers that have been marketed.” A photograph in that report shows a UBI 15 A two-pole breaker carrying 60 A without tripping, with a burn mark developing on the side. That is not a listing-mark argument. It is a trip-test argument.

Canadian Stab-Lok-type equipment is a related design under a different regulator. Federal Pioneer, later associated with Schneider Electric Canada, used the Stab-Lok architecture. Aronstein’s 2024 revision added enough Canadian samples, he writes, to conclude with statistical certainty that those breakers also have a high defect rate and do not provide the circuit protection the Canadian electrical code requires. That is an engineer’s finding about a related product. It is not a US recall.

What Canada actually recalled is narrower, and it is worth quoting so the two stories stay apart. Ontario’s Electrical Safety Authority, RCL97-05 (originally 1997, revised December 2017), records a Schneider Canada voluntary replacement of Federal Pioneer NC015 and NC015CP single-pole 15 A Stab-Lok breakers manufactured between 1 August 1996 and 11 June 1997. The hazard named is “breakers MAY NOT TRIP.” Identification is a square or rectangular blue handle. Replacement units have a hole drilled in the handle, or, after 1 January 1999, a rounded ribbed blue handle; black-handle breakers were outside the programme. That is two 15 A models, a ten-month production window, and a Canadian authority. It does not apply to a US FPE panel installed in 1968.

Challenger needs a sentence of its own, because the name sits on both sides of this distinction. After FPE, the residential line passed through Challenger Electric. Some Challenger-labelled breakers were made for Stab-Lok buses. A Challenger panel — Type C full-width, Type A half-space side-clip — is a different family. The door label and the breaker type marking tell you which you have. Treating every Challenger door as an FPE recall is how people pay for the wrong job.

Permit, grounding, AFCI and GFCI — what the adopted code actually adds

NFPA 70, the National Electrical Code, is a model code. States and cities adopt an edition, often with amendments, on their own calendar. The 2023 edition is the current published book as of this page; many jurisdictions still enforce 2017 or 2020, and a few have already moved. Nothing in this section declares a given house illegal. It describes what a permitted load-centre replacement commonly has to meet, and why the invoice is larger than a box of breakers.

The permit is the job becoming visible to the AHJ. A load-centre replacement is electrical work. The electrician pulls the permit in the jurisdiction that will inspect. A bid with $0 on the permit line is not evidence that the work is exempt; it is evidence that the bidder has not said who is pulling it. The inspection is what an insurer or a buyer’s inspector will ask to see. HyreElectrical does not pull permits.

Grounding, Article 250, is often the quiet half of the labour. Service equipment needs a grounding electrode system and a main bonding jumper. Houses that received FPE equipment in the 1950s through the 1970s were built to earlier editions. Typical field findings, none of which you can see with the dead-front on: a grounding electrode conductor landed on a water pipe that has since been interrupted by plastic; no supplemental rod; an undersized GEC; the neutral-ground bond at the wrong enclosure; no concrete-encased electrode even where one exists in the foundation. Opening the service equipment is when the AHJ typically requires that system to meet the adopted edition. That work is in the electrician’s scope. It is not a whole-house rewire.

Branch circuits without an equipment grounding conductor are a different finding. Replacing the panel does not pull a ground wire to the bedrooms. Ungrounded receptacles stay ungrounded unless someone adds an EGC under 250.130(C), installs GFCI protection with the “no equipment ground” marking, or rewires. Mixing those options onto a Stab-Lok invoice without naming them is how a load-centre swap becomes an unfinished argument six months later. Aluminium branch wiring, knob-and-tube, and cloth-insulated cable are likewise separate scopes; a new listed panel does not remediate the terminations in the rooms. See whole-house rewiring when the walls are the other half of the problem.

AFCI, 210.12, is the edition your jurisdiction adopted. Combination-type arc-fault protection for 120-volt, 15- and 20-ampere dwelling circuits has expanded from bedrooms (1999/2002) through most habitable rooms (2014 and later). Section 210.12(D) requires AFCI when existing circuits are modified, replaced or extended in those areas, with an exception — in editions that carry it — where the extension of existing conductors is not more than 6 feet and no additional outlets or devices are added, not counting the panel, junction boxes, or similar. Some AHJs treat a full panel replacement as triggering AFCI on the relocated circuits; some apply the 6-foot exception and leave existing runs on standard breakers. The electrician asks the department that will inspect, in the adopted edition, before the quote is compared to another bidder’s. This page will not declare which way your city goes.

GFCI, 210.8, is the same local question with a different list. Dwelling-unit GFCI locations have expanded across editions: bathrooms, garages, outdoors, crawl spaces, unfinished basements, kitchens, laundry, and, in later editions, 125- through 250-volt receptacles in those spaces rather than 125-volt only. Kitchen and laundry circuits often need both AFCI and GFCI; a dual-function breaker at the new panel is how that is commonly delivered. A 1970s FPE panel will not have those breakers. A like-for-like count of “one 20 A for the kitchen, one 20 A for the laundry” is therefore not a complete bill of materials under a current adopted edition.

Split-bus and 230.85. If the FPE is a rule-of-six panel with no main, the replacement typically includes a main breaker. If the service itself is being changed, the 2020 and later editions require an outdoor emergency disconnect under 230.85 — a meter-main, or a separate enclosure. Whether that SKU is in this job depends on whether the service is in scope and on the adopted edition. It is a reason the indoor load-centre swap and the service upgrade must be named separately on the quote.

The panel-space checker is useful after the AFCI/GFCI count is known; it does not size the new panel. The cost page is where those variables turn into money. This page will not invent a national average for either.

Insurance: what we can cite, and what we will not

Contractor pages often say “insurers will not cover FPE.” We did not find a state department-of-insurance bulletin or a FAIR-plan document that names Federal Pacific Stab-Lok as a coverage exclusion. Underwriting is company policy, not a national rule, and it changes. A blog’s “uninsurable” is not your declarations page.

What we can cite, as named carriers’ loss-control positions:

Cincinnati Insurance, in a 10 July 2025 loss-control article, recommends that the entire FPE panel be replaced with a modern panel by a licensed electrician, and treats FPE bolt-on equipment the same as Stab-Lok for that recommendation. The article describes field identifiers — the Stab-Lok word, red-painted handles, the white numbering strip with rounded corners and tick marks — and notes that the 2,800-fire / 13-death / $40-million figures in circulation “focus only on residential fires” and that “the actual numbers are impossible to know.” That last sentence is the honest one. Cincinnati is citing modelled estimates, not a CPSC census. The article is advisory loss-control, not an underwriting bulletin and not a department-of-insurance rule.

The Hanover Insurance Group, in a risk-solutions note carrying Hartford Steam Boiler copyright 2014 and an LC February 2019 code (171-1140), states that CPSC investigated Stab-Lok breakers failing UL tests, “did not have the funding to further investigate this problem or arrive at a definitive conclusion,” and that tests by CPSC and independent engineers found certain Stab-Lok breakers do not trip to UL requirements and can jam in the on position, with overheating at panelboard bus connections. Hanover’s published recommendation: when a Stab-Lok installation is discovered, “the safest course of action is to replace it with a completely new panelboard and breaker installation.” The same note cites the New Jersey 2002 partial summary judgment by docket number. Hanover’s own disclaimer says the material does not guarantee loss prevention and does not assume a duty. Read it as named-carrier loss control, not as a statute.

Recommendation: if a carrier, a broker or a home inspector has raised the panel, get the requirement in writing — brand, what must be replaced, whether a permit and a final inspection are part of the condition, and on what deadline. Then have a licensed electrician quote that scope. Two carriers’ loss-control pages pointing the same way is a reason to take the conversation seriously. It is not a federal recall, and it is not proof that every admitted insurer in every state will cancel. Ask the company that will actually write the policy.

Cost belongs on the cost page

HyreElectrical does not publish a national average for FPE replacement, because we do not have a measured job-cost dataset and will not invent one. The variables that move the invoice are the same as any panel upgrade: whether the service is changing, the run, AFCI/GFCI count under the adopted code, whether the enclosure can stay, the grounding electrode work, and the permit. A Stab-Lok replacement that is a like-for-like load-centre swap on an existing 100 A service is a different job from a 200 A service change.

Ask every bidder for the same written scope: confirm the panel is FPE Stab-Lok, whether the enclosure is being replaced, who pulls the permit, whether the utility is involved, what happens to the grounding electrode system, which AFCI/GFCI the adopted code is being read to require, and what happens if the branch wiring is a separate defect. Three numbers that do not share that scope cannot be compared.