Zinsco and FPE Panels: Hazard or Sales Scare

The line in the inspection report is short, and it is not an opinion about your house: Federal Pacific Stab-Lok panel noted. Recommend evaluation by a licensed electrician. Or single strand aluminum branch wiring observed. Or Zinsco/Sylvania panel. Then the phone call happens, and somewhere in it the words recalled, uninsurable and known fire hazard turn up, usually attached to one price for replacing everything.

Those three names travel together in the sales conversation. They do not travel together in the record. One has a thick federal file with two named repair products in it. One has a two-year federal investigation that ended in a shrug, plus a fraud finding in a state court two decades later. One has almost nothing on paper at all. Same recommendation on the report; three completely different evidentiary situations underneath it.

A caveat that matters more here than in most pieces on this site: nothing below is a judgment about your panel, because judging a panel means standing in front of it with the cover off, and I am not licensed to do that. A heat-damaged bus bar is a heat-damaged bus bar whatever any document says. What follows is the paperwork behind the three claims, read on 29 August 2026, and which lines in a quote each one does and does not justify.

The aluminum file is the thick one, and it names products

CPSC's Publication 516, Repairing Aluminum Wiring, dated June 2011, is the document everything else quotes. Its headline finding comes from a national survey run for CPSC by the Franklin Research Institute: homes built before 1972 and wired with aluminum are 55 times more likely to have one or more wire connections at outlets reach "Fire Hazard Conditions" than homes wired with copper. The publication defines that phrase rather than leaving it atmospheric — cover plate mounting screws reaching 149 °C (300 °F), or sparks emitted from the receptacle, or charring of the material around it.

CPSC also fences its own number in, twice, in the same paragraph. The survey covered only the connections at outlets; it did not address the other aluminum connections and splices in a house that are also prone to fail. And no information was developed for aluminum-wired homes built after 1972.

Read the recommendation itself before reading the rest of this as a debunking. Where a home has aluminum branch wiring that has not been remediated, CPSC recommends a permanent repair of the connections. The document is also blunt that you cannot wait for symptoms: failing aluminum connections seldom give easily detected warning signs, and have been reported to fail and overheat with no prior indication. Nothing below is an argument that the wiring does not matter. It is an argument about which repair, on how many connections, at whose price.

Whether the recommendation is even about your house turns on a date and a marking. Aluminum branch circuits went in roughly between 1965 and the mid-1970s, and homes built before 1965 are unlikely to have them. A jacket marking reading CU-clad or Copper-clad alongside the Al means copper-coated aluminum, which the publication expressly excludes from its repair recommendations.

The part quotes get wrong most often is the list of permanent repairs. There are three, and CPSC states the test each has to pass — safe, effective, permanent:

1. Complete replacement with copper cable. Called the best method available, and in the next breath called impractical or prohibitively expensive for many homes.
2. COPALUM. A copper pigtail joined to the aluminum with a crimp sleeve, installed only with a dedicated power tool that compresses the connector with upwards of 10,000 pounds of force, and only by installers the manufacturer has trained.
3. AlumiConn. A setscrew connector, named as the next best alternative where COPALUM is unavailable in your area, and described as too new to have COPALUM's long-term field history.

Everything else sits under a heading called recommendations against using other repair methods. Other crimp connectors: not evaluated for connecting copper and aluminum, strongly recommended against. Twist-on pigtails: in CPSC-sponsored laboratory and life testing, substantial numbers overheated severely, and the staff writes that it is possible some such repairs "may be prone to even more failures than the original aluminum wire connectors." CO/ALR receptacles and switches: better than ordinary devices, but they failed in laboratory tests under conditions that simulated actual use with no overstress, they do not exist for ceiling fixtures and permanently wired appliances, and they are "at best, an incomplete repair." Both rejected methods appear once more at the end, as an emergency or temporary repair for a single failed aluminum termination, followed by a recommendation to do the complete repair as soon as possible.

Two scope limits are worth reading before you accept a whole-house number. The repair covers every aluminum connection and splice in the branch circuit system, explicitly including junction boxes and directly connected appliances such as dishwashers and furnaces — not only the receptacles a bid tends to count. And it does not apply to the larger gauge aluminum used for service entrance cable, ranges and dryers. A line item proposing to replace your aluminum service entrance conductors on the strength of Publication 516 is citing a document that says the opposite.

One practical note on availability, and its limits. The COPALUM web address printed in the 2011 publication, tooling.tycoelectronics.com, no longer resolves in DNS. The same page also prints a phone number and a postal address for the manufacturer's aluminum wire repair programme, and the manufacturer's current site did not open on the day this was checked, so whether the programme still trains and supplies installers in your area is not something the publication can tell you and not something I established. That is a question for the contractor, along with the connector brand written into the quote by name and whatever training or authorisation the installer holds for it. The publication is explicit that only manufacturer-trained electricians are authorised to install the crimp.

What the 1983 Stab-Lok release says, and what it carefully avoids saying

Press release 83-008, Commission Closes Investigation of FPE Circuit Breakers, issued 3 March 1983 and revised 18 February 2011, takes four minutes to read and is worth all four, because both sides of the argument quote half of it.

The investigation began in June 1980, when Reliance Electric — an Exxon subsidiary and FPE's parent — reported to the Commission that many FPE circuit breakers did not fully comply with Underwriters Laboratories requirements. Commission testing then confirmed that these breakers fail certain UL calibration test requirements. That is the agency's own sentence, not a plaintiff's. It was not an informal look, either: a full-text search of cpsc.gov turns up Docket No. 81-7, In the Investigation of Federal Pacific Electric Company and Reliance Electric Company, an authorisation of subpoena and special order in the same matter, and Commission ballot decisions from September and December 1982.

Then the other half, which the sales call never reaches. The Commission was closing the file because "the data currently available to the Commission does not establish that the circuit breakers pose a serious risk of injury to consumers." It was unable at this time to link the calibration failures to the development of a hazardous situation. The manufacturer's position was that at the few overload levels where its breakers might fail to trip in realistic use, currents would be too low to generate hazardous temperatures in household wiring. Staff had insufficient data to accept or refute that, and put the cost of finding out at several million dollars against a total agency budget of $34 million for fiscal 1983. So it closed, while saying it could reopen if further information warranted.

The document reaches less far than either side of the phone call needs it to. The investigation focused primarily on two-pole breakers manufactured before Reliance acquired FPE in 1979 — it is not a verdict on every breaker in the panel. And the 2011 revision adds a note in the agency's own voice: the matter was closed without making a determination as to the safety of FPE circuit breakers or the accuracy of the manufacturer's position.

No recall followed, then or since. Nor any finding that the breakers are fine. Both of the confident sentences a homeowner hears on the phone are unsupported by this document.

The record did not stop in 1983. It moved venues. In a New Jersey class action, a partial summary judgment dated 15 August 2002, Judge Bryan D. Garruto, Superior Court of New Jersey, Law Division, Middlesex County, Docket L-2904-97, found the company had committed fraud under that state's consumer protection statute for applying testing-laboratory labels to breakers that did not actually comply with the standard. A quarterly report the parent company filed with the Securities and Exchange Commission states that listings had been obtained through "deceptive and improper practices," that most of the company's circuit protective products lost that listing afterwards, and that the company was working to correct the deficiencies and regain them.

Both citations come from the reference list of Jesse Aronstein's peer-reviewed Faulty Residential Circuit Breakers — A Persistent Fire Safety Problem, IEEE Open Journal of Industry Applications vol. 4 (2023), 75–86, which is open access and free to read. Two things to know before you open it. The paper never prints the brand: IEEE editorial rules restrict the use of actual brand names, so FPE appears throughout as "Brand X," identifiable from the documents cited for it. And the paper dates that SEC filing to 1982 in its text and to the quarter ending 31 March 1983 in its reference list, which is the sort of discrepancy worth knowing about before you quote a date at a contractor.

For scale, that paper carries forward an estimate from Aronstein and Lowry's Estimating Fire Losses Associated With Circuit Breaker Malfunction, IEEE Transactions on Industry Applications vol. 48 no. 1 (2012): about 2,800 residential fires a year associated with defective operation of these breakers, with roughly 116 injuries, 13 deaths and $40 million in property damage. Set that beside NFPA's estimate of 32,620 home fires a year involving electrical distribution and lighting equipment, causing 430 deaths a year, averaged over 2015–2019 and published February 2022. The claim is that one discontinued breaker line accounts for something under a tenth of that category of fire and about three percent of its deaths. Serious. Also not the same sentence as "this panel is going to burn your house down."

Where the numbers come from matters as much as the numbers. The agency did generate its own data: CPSC retained Wright-Malta Corp. in 1980 to test 122 full-size two-pole Stab-Lok breakers, new and used, and 51 percent of them failed to trip as required at 135 percent of rating, with 10 percent of the sample jamming on one pole. UL testing reported to an industry committee in 1979 found 5 of 55 breakers of various brands failing the same requirement. Both of those figures reach the public through the 2026 paper below rather than through an agency publication. Beyond them, essentially every published FPE calibration figure comes from one independent testing project run by the author above, who has been testing residential breakers since 1981 and publishes running totals at fpe-info.org — currently 2,615 breakers of the brand at a 23 percent calibration defect rate. His own biography in that journal places him at Wright-Malta from 1974 to 1984, so the agency's contractor data and the independent project share an institutional origin. The work is peer reviewed, the counts are public, no contradicting data has been published, and it is still a narrow evidence base. You are entitled to know that when a contractor says "the studies."

Zinsco's folder is nearly empty, and that cuts both ways

Here is the search I ran and what it returned, so you can repeat it. CPSC's own site search, set to include archived material and to cover every content type — recalls, press releases, Records of Commission Action, adjudicative dockets — returns no results at all for Zinsco, on 29 August 2026. The same search for Federal Pacific Electric returns eleven documents, including the docket and the 1983 release. The same search for Sylvania returns recall notices — for lamps and light kits, none for panels — which tells you the index does hold recalls and would surface one if it existed.

Say plainly what that does and does not establish. A site search is not the same instrument as a subpoena or a docket, and the agency's dedicated recall search page does not accept a search term through its address at all, so this is one route to the same index rather than two. And the published breaker-testing literature identifies brands by number, not by name, precisely because journal rules restrict brand names — so I cannot tell you that no tested set was Zinsco. What I can tell you is that no published calibration series is identified as Zinsco. What exists under the name is home-inspection and forensic material: photographs of burned breaker-to-bus contact points, and the observation that the bus erodes where the breaker clips onto it.

So a quote asserting that Zinsco panels were recalled asserts something that is not in the agency's index, and a report claiming the failure rate has been quantified goes further than anything published under that name.

None of which is a clean bill of health, and it should not be read as one. An empty file means nobody with subpoena power ever finished looking. That is a different thing from looking and finding nothing, and the absence of a federal document is not evidence that a panel is safe. The questions that do have answers for a Zinsco are narrower, and all three are answered by somebody other than you: what the inspector's photographs of the bus and the breaker contacts show once the panel cover is off, whether breakers you can actually buy are listed for that enclosure, and whether your carrier will write the policy.

The only breaker test that means anything runs on a bench

A residential branch circuit breaker must trip at or below 135 percent of rated current — 27 amps on a 20-amp breaker. The calibration test applies that current and waits, up to an hour.

That number matters because of a specific sales line. In Ineffective Calibration Testing of Residential Branch Circuit Breakers, IEEE Open Journal of Industry Applications vol. 7 (2026), 389–395, 100 breakers already known to be defective — three brands, manufactured between about 2006 and 2022, all failing to trip at 135 percent — were retested. Every one passed the UL 489 calibration test at 200 percent of rating, which requires a trip within 120 seconds, and every one also passed the NEMA AB 4 test at 300 percent, which allows 50 seconds. A 100 percent error rate for detecting the defect that matters. Note the manufacturing dates: this is not only an old-panel problem, and every one of those breakers carried a testing laboratory's listing mark.

Which means the little test button proves the mechanism moves, and a driveway demonstration at high current proves nothing about calibration. If somebody offers to test your breakers, there are only three questions: at what percentage of rating, for how long, and does the breaker come out of the panel to do it.

The same literature complicates the cheaper counter-offer, without settling it. Newly manufactured Stab-Lok-compatible breakers are sold today; Connecticut Electric lists its replacement Federal Pacific breakers as newly manufactured and Intertek ETL listed. Against that, the 2026 paper reports calibration results for 622 breakers "that are the same design as Brand X, but marketed under different Brand names," at a combined 38 percent failure rate. The paper does not say which brand names those 622 were, or when they were made, so it is not a test of the box you would buy this week — it is a reason to ask for that manufacturer's own 135 percent test data before treating a breaker swap as the cheap way out. It is a second purchase with the same question attached to it.

The sentence that actually moves money is on the insurer's form

Underwriting, not code, is usually what forces the decision, so read what an insurer's own form asks for. Florida's state-created insurer publishes its 4-Point Inspection Form on a public page; the current sample is Insp4pt 03 25, and the March 2025 bulletin announcing it says the previous versions were deleted from the site, so an inspector still working from an older printout is working from a form the insurer has withdrawn.

The electrical section asks for panel brand and model, panel age, total amps and whether the amperage is sufficient for current usage, and it requires photographs of the main service panel with the interior door label and of the electrical box with the panel off. Under branch circuit aluminum wiring it asks the inspector to describe the usage of all aluminum wiring, then gives exactly two tick boxes for remediation: Connections repaired via COPALUM crimp and Connections repaired via AlumiConn. A note at the top of the section says separate documentation of any single strand aluminum wiring remediation must be provided and certified by a licensed electrician — and that sentence is one of the changes the insurer listed in its March 2025 bulletin, which describes it as a clarification to the electrical system section. The whole form has to be completed, signed and dated by a verifiable Florida-licensed professional.

That layout has consequences in both directions. The two CPSC-approved pigtail methods are the two the form recognises, so the cheap twist-on version has no box to tick and can cost you the policy instead of saving you money. And there is no tick box anywhere on that form for FPE or Zinsco: it asks for brand and model, then leaves the judgment to underwriting. Which is why "your insurance company requires this" is a claim to verify with your insurance company, in writing, before you buy the remedy — and why the licence, bond and complaint record of whoever signs that certification is worth ten minutes.

Which lines in the quote are carrying weight

For an aluminum remediation bid, the quantity that sets the price is connections, not rooms:

• the connector by brand name, and the installer's authorisation if the answer is COPALUM
• a count of connections and a per-connection unit price, covering every device, junction box, ceiling fixture and directly connected appliance, because the CPSC recommendation covers all of them
• what happens at connections that turn out to be inaccessible, and at what price
• the signed remediation letter your carrier's form is going to ask for
• permit and inspection: Publication 516 closes by saying all modifications and additions to installed wiring should be performed and inspected in accordance with local regulations

For a panel replacement bid, the traps sit elsewhere:

• brand, model and rating of the new panel, and the main breaker rating, which is not the same number — that distinction is the whole of the load calculation this site walks through, and it decides whether you are buying capacity or condition
• service conductors, meter socket and mast: in scope or not, and who schedules the utility disconnect
• AFCI and GFCI breakers, how many and at what unit price, since a replaced panel often has to be populated to current local rules and those breakers are not cheap
• outage duration, and a labelled circuit directory at the end of it
• the permit line, priced rather than waived — what unpermitted work leaves behind surfaces at closing, and a panel swap is the least concealable work in the house

What a real recall looks like, for comparison

On 16 June 2022 CPSC announced recall 22-159: Schneider Electric recalled about 1.4 million Square D QO Plug-On Neutral load centers manufactured between February 2020 and January 2022, date codes 200561 through 220233, for overheating that poses thermal burn and fire hazards. Remedy: repair or replace, at the manufacturer's expense. Reported incidents at announcement: one loose wire, no injuries.

That is the anatomy of the word. A recall number, a date code range tied to identifiable equipment — printed inside the cover on outdoor units, and, the notice says plainly, not visible to the homeowner on indoor ones — a named remedy, and somebody other than you paying for it. One reported incident produced all of that in 2022, which is worth holding next to a 1980s file that produced none of it.

So the word in your quote either has a number attached or it does not. If it does, look it up in CPSC's recall database and read what the manufacturer owes you. If it does not, it is doing sales work — which settles what the word proves, and nothing about what is inside the enclosure. The decision in front of you is duller than the phone call made it sound: what the inspector's photographs from inside the panel show, what your carrier will put in writing, and whether the arithmetic on the alternatives — pigtails, a subpanel, breakers, nothing at all — comes out under the number on the page. If the photographs show heat damage, none of the paperwork above argues with them.