Keeping Aquariums & Reptiles Alive in a Power Outage

SnuggleSafe, UniHeat, Aquarium Co-Op, Penn-Plax, and Aquatop are trademarks of their respective owners; EmergencyPetPrep is not affiliated with or endorsed by any of them.

A dog or cat can usually ride out a power outage on blankets and body heat. A 40-gallon reef tank or a bearded dragon’s basking lamp can’t. Fish depend on a heater and, eventually, oxygenation; reptiles depend on an external heat source just to digest and move normally. Both start losing ground the moment the grid does, and almost nobody’s evacuation kit accounts for them. Below is what to do in the first minutes, what to do if the outage stretches past a few hours, and which gear actually has a published spec behind it, including where the specs simply don’t exist yet.

The American Red Cross’s own pet-disaster guidance is dog-and-cat focused, and explicitly refers reptile and exotic-pet owners elsewhere for species-specific guidance. That’s not a knock on the Red Cross. It’s a real gap, and it’s the gap this page is built to fill.

What This Page Answers, and What It Hands Off

This is the generalist playbook for everything in the house that lives in a tank or a terrarium: the first hours, the clock and what moves it, the gear whose specs are actually published, and the check to run once the grid returns. It is deliberately not a species manual, and the reason is visible in the tolerance table further down. None of the duration figures in it trace to a vet-school or clinical source: the bearded dragon husbandry range comes from VCA, and the ball python and leopard gecko figures come from care guides, but no primary source publishes an hours-without-heat ceiling for any of the three. There is no cross-species clinical table to average them into, and a hub that flattened them into one number would read more confidently and be less true.

The question you arrived with	Where the answer honestly sits
What do I do in the next hour, for any tank or terrarium in the house	Here, in the next two sections
How many hours do I have	Here, as a staged plan, the four inputs that move it, and a timer you run on two thermometer readings of your own tank rather than on a formula nobody has published
Once the timer runs out, what do I try, and in what order	Here, in NC State’s own published order of escalation, from restoring power to the one step it labels an absolute last resort
What is the safe floor for my particular species	Your species’ own husbandry source, read against the vet-emergency threshold on this page, which overrides every species figure below it
How long will this heat pack last in this enclosure	Partly here: this page gives you the manufacturer’s published duration and the insulation rule. Our heat pack runtime page carries every published figure and sets out why no publisher converts a rated duration into hours for a specific enclosure, so treat the rating as a property of the pack rather than as arithmetic you can run on your volume and your room temperature
If I can only power one thing on a reef or saltwater tank, what is it	Our reef tank power outage priority guide, which carries the published evidence on flow versus light, the two peer-reviewed clocks behind it, and the manufacturer runtime figures for a battery backup on a direct-current pump. This page carries the rest of the reef picture: the failure chain, the heater wattage problem, both directions of the temperature problem, and salinity creep
What order do I bring a reef system back up in	Not here. The restoration rule that is time-critical for a reef system is on this page, in the salinity section. The staged timeline table is scoped to freshwater, so this page does not hand you its rows as a reef restart. The full media-and-cycle restart is a longer sequence than an outage playbook should carry
Why did the heater not come back on when the power did	Here, in the restoration section near the end. It is the failure owners are least prepared for

Two rules hold the rest of the page together. Symptoms override clocks, every time. And any figure below that is not clinical is labelled as not clinical, because in this category the most confident numbers are usually the least sourced.

Do This First: The Outage Just Started

Before you touch a heat pack or order anything, do the free things.

• Don’t feed fish or reptiles. For fish, digestion consumes oxygen you can’t spare in a tank that’s about to lose aeration, per NC State veterinary aquatic medicine faculty. For reptiles, food can’t be properly digested at a lowered body temperature and risks bacterial overgrowth.
• Disable auto-feeders on aquariums so uneaten food doesn’t foul the water while you’re not watching it closely.
• Keep the tank lid closed to slow heat loss, and move the tank away from drafty windows or doors if you can do it without jostling it.
• Check what’s actually failed. A dead heater in a warm room is a different problem than a dead heater in a cold house in January. Room temperature drives your timeline as much as the tank or terrarium’s own equipment does.
• Do not open the enclosure repeatedly “to check.” Every time you lift a tank lid or a terrarium screen top, you let out the heat you’re trying to keep in.

For the first 1–2 hours in a freshwater tank, Aquarium Co-Op’s guidance is simple: do nothing else. Aquarium Co-Op files that article under its Freshwater Aquarium Blog, and this page does not extend the opening wait to a saltwater or reef system: that case is the reef section further down, and the question of what to power first is on our reef tank power outage priority guide. NC State’s Gregory Lewbart gives two different windows rather than one, and the difference matters. Tropical and temperate fish tolerate water temperatures in the high 80s or low 90s°F “for a day or more.” Most tropical fish tolerate the low 60s or even high 50s°F “for several hours.” You have longer on the warm side than on the cold side, and you do not need to panic-buy a battery pump in the first thirty minutes.

Freshwater Fish and Aquariums: A Staged Timeline

Here’s the escalation path for a freshwater tank, built from NC State’s veterinary guidance and Aquarium Co-Op’s staged-response plan. Nothing here is a lab-tested countdown. It’s the best publicly available guidance we could source, and we say so at every step.

Time since power loss	What to do	Source
0–2 hours	Nothing, in a freshwater tank. No feeding, no extra aeration, no water changes.	Aquarium Co-Op
2–8 hours	Monitor water temperature if you have a battery or battery-backup thermometer. Keep the lid closed.	Aquarium Co-Op
8 hours+	Run supplemental aeration (battery air pump, or manual agitation with a cup) for about 1 hour, then repeat roughly every 8 hours. In heavily stocked tanks, repeat closer to every 4 hours.	Aquarium Co-Op
Any point	If ammonia tests above 1 ppm, do a 10–30% water change.	NC State (Gregory Lewbart)
After power returns	Withhold feeding for 24 hours to let biofiltration catch back up.	Aquarium Co-Op

The scope of that table, stated at the table rather than left to the reef section further down: every row is for a freshwater tank. This site does not extend the 0 to 2 hour do-nothing opening to a saltwater or reef system. Two more things travel with the table. Aquarium Co-Op’s own exception outranks its own intervals: if you notice your fish gasping for air, you may need to add oxygen earlier and increase how often you run it. And if what you have is a reef or saltwater tank, the question of what to power first is on our reef tank power outage priority guide, which reads the published evidence as pointing at water movement first rather than at a waiting window; the reef failure chain, the heater wattage problem, both directions of the temperature problem and salinity creep stay here, further down this page.

On temperature specifically, NC State’s guidance gives you real margin: most tropical fish tolerate the high 80s to low 90s°F, or the high 50s to low 60s°F, for a day or more on the warm side, and several hours on the cold side, without lasting harm. The bigger threat in a longer outage isn’t the thermometer, it’s falling dissolved oxygen, which is exactly why the staged aeration plan above exists.

The Four Things That Move Your Clock

That table is a default, not a countdown. Four things decide whether your tank runs ahead of it or behind it, and no source publishes a formula that combines them, so what follows is the direction each one pushes and the source that establishes the direction. Anyone selling you an exact hours-remaining figure for your tank is doing arithmetic nobody has validated.

Input	Which way it moves your clock	The sourced reason
Water volume	More water buys more time	Aquarium Co-Op, plainly: “greater volumes of water do not change temperature as easily.” A small tank in a cooling house tracks the room; a large one lags behind it.
Stock load	A heavier bioload spends the clock faster	Aquarium Co-Op’s own exception to its 8-hour aeration interval is the heavily stocked tank, which it says may need aeration every 4 hours instead. Oxygen, not temperature, is what the load consumes.
Room temperature	The room is both the ceiling and the floor	NC State’s Gregory Lewbart: most tropical fish tolerate the low 60s or even high 50s°F for several hours, and the intervention line is “Once temperatures dip to the mid-50s”. In a house that holds its heat, a freshwater tank may never reach that line at all.
Insulation	It slows the loss. It never adds heat	NC State: “insulate your aquarium with a blanket, sleeping bag or newspapers.” Aquarium Co-Op supplies the mechanism, which is why the technique matters: use multiple layers, because the air trapped between them is what insulates.

Read them together rather than one at a time. A lightly stocked large tank under three blankets in a house that is still warm is a different emergency from a heavily stocked small tank in an unheated garage, and the staged table treats the two identically until you apply these four.

There is one more figure worth having before you start fixing anything, because it caps how fast you are allowed to correct. NC State puts a safe rate of change this way: “A 10 percent water temperature change every hour or two should not endanger fish.” That is a ceiling on the rescue, not only on the outage. Pouring hot water into a chilled tank to catch up is its own injury, and it is the mistake a panicking owner makes at hour ten.

For reptiles the same clock splits into two, and owners routinely run them together. One clock is how long the animal is safe at the temperature the room is drifting toward, which is a husbandry-and-symptoms question and is answered in the reptile sections below. The other is how much usable heat the pack, bottle or charged pad has left, and that one has no arithmetic behind it: the rated hours are a property of the pack rather than of your enclosure, and our heat pack runtime page sets out, source by source, why none of the publishers converts a rating into hours for a given volume, insulation and room temperature. Run the second clock on the manufacturer’s rating and a thermometer together, and treat anyone who hands you a computed hours-remaining figure for your enclosure as having built the model themselves.

Your Own Survival Timer: Measure the Rate, Don’t Model It

The four inputs above tell you which direction your tank sits off the default. They don’t tell you when. There is a way to get an actual hours figure, and it isn’t a formula: measure the rate your own tank is losing, because that measurement already contains the volume, the insulation and the room. Nobody has to model what you can read off a thermometer, which is why this page gives you a procedure instead of a lookup table.

1. Take two readings, an hour apart. Water temperature and the clock time, then water temperature and the clock time again. Note the room’s temperature beside each one, because the room is what the tank is drifting toward.

2. Subtract. The difference is degrees per hour for this tank, in this room, under whatever you have wrapped around it. That number does not exist anywhere on the internet, because it is a property of your house.

3. Find the line you are heading for. For tropical fish, NC State publishes the intervention line: “Once temperatures dip to the mid-50s” is where Gregory Lewbart’s guidance switches from monitoring to raising the temperature. If you keep temperate species instead, goldfish or koi, NC State says they “should be fine without changing the environment.” So the cold half of this timer is not your emergency. In a summer outage the ceiling replaces the floor, and NC State puts tropical and temperate tolerance in the high 80s or low 90s°F “for a day or more.”

4. Divide. Degrees of gap divided by degrees per hour is your hours to the intervention point. Purely to show the arithmetic, with numbers chosen at random and meaning nothing about your tank: a thermometer reading 74°F and then 72°F an hour later is losing 2°F an hour, the gap down to the mid-50s is roughly 17°F, and the timer reads about eight hours. Your two readings are the only ones that mean anything.

One bound rides along with that number, and the division does not carry it on its own. NC State’s tolerance for the low 60s and high 50s runs “for several hours” rather than indefinitely, so the hours your tank spends drifting down through that band are already being spent against the same budget. A slow-cooling tank is where this bites: a large insulated tank losing half a degree an hour from 68°F divides out to a 26-hour timer, and roughly half of those hours are spent inside a band the source calls tolerable only for several hours. Read the result as the latest hour you can act, not as permission to wait for it, and the slower your measured rate, the more the reading overstates how long you actually have.

5. Read it again in an hour or two. This is a measurement you repeat, not a figure you compute once, and it describes only the room you have right now. If the house is still cooling, nothing about the rate you just measured is settled. If the two readings come back the same, you don’t have a temperature emergency at all, and the second clock below is the one that matters.

Where stock load went. It is the one input of the four your thermometer cannot see, because stock load does not move the temperature. It moves the oxygen, and oxygen is a second clock that is not measured in degrees.

The oxygen clock. It runs on the staged intervals in the table above rather than on a thermometer: Aquarium Co-Op’s roughly one hour of aeration every 8 hours once you pass the 8-hour mark, tightened to every 4 hours in a heavily stocked tank. There is also an observable that outranks both intervals, and it is the one to memorize, because it arrives without warning you: Aquarium Co-Op says that if you “notice that your fish are gasping for air”, you may need to add oxygen earlier and increase how often you run it. Fish at the surface are not waiting for your interval to come around.

Terrariums run the same procedure and stop at a different line. Two readings an hour apart give you degrees per hour for an enclosure just as they do for a tank. What this page will not hand you is the number to divide toward, and the reptile sections below explain why: the bearded dragon’s husbandry range has a vet-reviewed source (VCA) behind it but no published emergency-tolerance figure, the snake and gecko figures come from care guides instead of a clinical source, and averaging any of those into one cross-species floor would read more confidently than any of its parts. Divide toward your own species’ published minimum, from your own species’ source.

When the Timer Runs Out: What NC State Lists, in the Order It Lists Them

Most outage advice hands you every option at once. NC State prints a list, introduced with the words “elevate the temperature using the following options” and a colon. It ranks exactly one item, the last, which it opens with the words “An absolute last resort”. Below are the six of Lewbart’s options that apply to an indoor tank, kept in the order the source prints them, with our notes attached rather than folded into them. Two of the source’s eight are left out of the numbering: the too-warm and direct-sunlight item, which is a summer instruction and is folded into the two summer notes below this list, and the temperate-species and pond-ice item, whose first half is already used in step 3 of the timer above and whose second half is about ponds rather than tanks. Reading the printed order as an escalation, running from the intervention that changes nothing about the animal’s world to the one that changes everything, is our note rather than a structure NC State claims.

1. Restore power to the equipment you already own. NC State says: “Use an alternate power supply to run the heater and pump/filter.” That supply “could be a generator or creatively used extension cord to a power source.” A neighbor’s working outlet and a long cord can be the entire intervention, and it sits first because it costs the tank nothing.
2. An external heat source, with a carbon monoxide caveat this page will not soften. NC State’s wording is “Use an alternate or safe external heat source such as propane or kerosene heaters.” The load-bearing word in it is safe. We do not restate that item as our own recommendation, because burning fuel indoors raises the same carbon monoxide question the generator section below answers, and that section is where our position lives. The heat sources this page does name are chemical packs and insulation, in the next subsection.
3. Move the fish. Per NC State: “Move the fish to a warmer location; heavy-duty Ziplock bags work well.” It names a bucket, tub or large jar as the alternates. The packing ratio and the survival window are in the bag-and-bucket section below rather than repeated here.
4. Move the aquarium. NC State’s own words are “Moving the aquarium is an option, though challenging.” It then publishes the figure that tells you how challenging: “up to 70 percent of the water may be discarded (a gallon of water weighs 8 pounds).” That sentence turns a vague dread of moving a tank into arithmetic you can do standing in front of it, and it is the reason a full tank is a two-person job and a drained one may not be.
5. If the fish can’t be moved and the water is too cold, warm it slowly. NC State’s instruction is to “add warm dechlorinated water to the aquarium.” That is governed by the same ceiling as everywhere else on this page: “A 10 percent water temperature change every hour or two should not endanger fish.”
6. Last, and NC State labels it as such: don’t pool systems. NC State’s own last item reads “An absolute last resort: mixing fish from one aquarium or pond with fish from another aquatic system.” Its reason: doing so “greatly increases the risk of spreading infectious viral, bacterial, fungal and parasitic diseases (and could also lead to aggression between the animals).” The neighbor with power and a spare tank is a good idea right up to the moment the two populations share water.

Two more from the same source that outage checklists routinely drop. On a summer outage the temperature instruction inverts: rather than insulating, NC State says to clear the tank surface “to increase the area for gas-exchange efficiency.” It names “stirring water with a whisk” as the manual way to raise oxygen. And on capture, NC State’s own supply list explains why a net is not the thing you lift with, since “herding fish into a plastic bag using a net is a much safer way to capture fish than using a net on its own.” Its reason: “Nets can damage the protective mucus layer and sensitive epidermis.”

Heat Packs for Aquariums: The Math

If the room itself is getting cold (a winter outage, an unheated space), a chemical heat pack taped to the outside of the tank, never directly in the water, can slow heat loss. Aquarium Co-Op publishes a rule of thumb of one heat pack for every 20–30 gallons of water, adjusted as needed. That’s retailer guidance, not a lab-tested figure, so treat it as a starting point and check the tank temperature rather than trusting the ratio blindly.

Two limits ride along with that ratio, and both push toward fewer packs rather than more. UniHeat’s own 40-hour pack page tells shipping customers to use only one heat pack per shipment to avoid overheating, so the US page for the pack itself argues against stacking rather than for it. And the per-pack increments that do get published, UniHeat’s own included, are scoped to a closed insulated box of about one cubic foot rather than to a tank standing in a room, which is why our heat pack runtime page will not turn a pack count into a temperature. Start counting from Aquarium Co-Op’s ratio if you like, then settle it on your own thermometer rather than adding packs on the strength of the arithmetic.

UniHeat’s shipping heat packs are a reasonable source for this because the manufacturer actually publishes duration numbers: 20, 30, 40, 60, 72, 96, and 120-hour versions, with the 72-hour pack averaging roughly 70°F inside a standard 1 cubic-foot insulated box. Two things matter for safety here, both manufacturer-stated: these packs need oxygen to activate, so never seal one in an airtight container, and they’re built for shipping boxes, not direct animal or glass contact. Insulate with a towel layer; don’t tape it straight to bare glass or skin.

Evacuating Fish: The Bag-and-Bucket Numbers

If you have to leave and can’t take the tank, NC State’s own transport guidance gives a workable analogue: fish packed sparsely (about 5 inches of fish per gallon) in a bag with roughly 1/3 water and 2/3 air can survive at least 36 hours, provided the external temperature stays adequate. That’s a transport figure, not a “leave them in a cooler for two days” endorsement. It tells you fish can travel further than most owners assume, if you don’t overcrowd the bag. For the full step-by-step version, including what to grab and the realistic survival window, see our guide on evacuating and transporting fish during a hurricane.

Saltwater and Reef Tanks: What’s Different

Everything above assumes a freshwater tank. A saltwater or reef system fails faster, and in more ways, during an outage, because more of its life support runs on electricity at once: the return pump, the powerheads, the protein skimmer, and on many tanks an auto top-off, all on top of the heater. Bulk Reef Supply’s outage guidance is blunt about the chain reaction. Without water movement from pumps and powerheads, it says, “dissolved oxygen levels will drop rapidly, especially in tanks with a higher bioload.” Filtration systems, such as protein skimmers and reactors, “stop removing organic waste.” That, in its words, “allows ammonia and other harmful compounds to accumulate quickly.” A reef tank of the same gallonage as a freshwater community tank has less margin, not more, which is why the staged freshwater timeline above, and its 1 to 2 hour do-nothing opening in particular, is not this site’s answer for a reef system. What to power first, if you can only power one thing, is a separate question with its own published evidence behind it, and it is on our reef tank power outage priority guide rather than here.

Marine Heater Wattage, and Why “Battery Backup Heater” Is Mostly a Myth

Reef animals live in a narrow temperature band, so heater sizing matters more here than in a hardy freshwater setup. Aqueon’s published rule of thumb is the cleanest manufacturer number to size from: 5 watts per gallon for aquariums 55 gallons or smaller, and 3 watts per gallon for those over 60 gallons, plus a larger heater “or a second heater if your aquarium is in an especially cold room or is located on an exterior wall or near an outside door.” That last clause is the outage-relevant one, because a winter blackout turns your whole house into that cold room. Two honest limits: Aqueon publishes one aquarium heater guide, not a separate marine chart, so the wattage math is the same for salt and fresh; and many reef keepers run to the top of that range split across two heaters for redundancy, which is keeper practice, not an Aqueon requirement.

Here’s the part the phrase “saltwater aquarium battery backup heater” tends to gloss over: an aquarium heater is far too power-hungry for the kind of battery that runs an air pump. A backup air pump sips well under 2 watts, so it runs a long time on modest power: Penn-Plax publishes 72 hours or more on two D-cells for the B11 (SKU SAB11) depending on the batteries used, against up to 48 hours on its retail listing, and the rechargeable Aquarium Co-Op unit (1.2 W) runs 20 to 40 hours per charge. A reef heater, typically 100 to 200 watts at Aqueon’s 5 watts per gallon for a 20 to 40 gallon tank, is a different order of magnitude, so “battery backup” for a heater realistically means a portable power station or generator, not disposable cells, and even a large power station buys hours, not days. For most reef keepers the honest temperature plan is insulation and warm-water bottles as bridges, with a power station or generator only if the outage runs long. Wrap the tank in blankets, towels, or foam board, and, per Bulk Reef Supply, float sealed bottles of warm water in the tank to hold temperature (or frozen bottles if the risk is overheating in a summer outage).

Powerheads and the Protein Skimmer: Why Oxygen Falls Faster

In a reef tank, oxygen doesn’t come mainly from one air pump. It comes from constant surface agitation and, on skimmed systems, from the protein skimmer injecting air as it runs. An outage stops all of it at once, the return pump, the powerheads, and the skimmer, so gas exchange nearly halts while a dense population of fish, corals, and live rock keeps consuming oxygen. Bulk Reef Supply’s fix is the same low-tech chain freshwater keepers use, scaled up: keep multiple battery-powered air pumps on hand, and “manually stir the water every hour” if you run out of aeration. The battery air pumps covered in the next section work identically in saltwater, and here airflow, not just runtime, carries extra weight because the bioload competing for that oxygen is higher (the Aquatop BREZA’s 2.5 L/min is the airflow standout, even though it publishes no runtime).

Coral Thresholds: Temperature and Oxygen

This is where honest sourcing gets thin, and the page will say so. There is no primary-authority “corals die at X°F after Y hours” table the way there’s a clinical reptile-hypothermia threshold further down this page. What is well supported: reef keepers hold temperature in a narrow band, broadly 76 to 80°F, with reef-keeping consensus treating roughly 83 to 84°F as the upper edge you don’t cross. Bulk Reef Supply’s own warning about what happens when the heaters and chillers stop is short: “Even slight fluctuations can stress or kill sensitive reef life.” That is why its entire reef-outage protocol is built around preventing the swing rather than reacting to it. Corals, especially small-polyp stony (SPS) species, tolerate far less temperature drift than a hardy freshwater fish, and they’re also among the first to suffer when dissolved oxygen falls. So the practical takeaway is the inverse of the freshwater timeline above: where a goldfish tank buys you hours, a reef tank’s honest answer is act sooner, aim to hold the temperature you already had, and treat falling oxygen as the real clock. Anyone quoting you an exact reef death-temperature with a stopwatch is going past what the sources actually support.

Salinity Creep on a Long Outage

One reef-specific problem freshwater keepers never face: salinity rises the longer the power’s off. Water evaporates but salt doesn’t, so, as Bulk Reef Supply puts it, “During evaporation, only gaseous H2O molecules leave the aquarium; all of the dissolved salts and minerals in the saltwater stay behind.” What is left concentrates. An auto top-off normally replaces that evaporated water with fresh RO/DI to hold salinity steady, but an ATO is itself a powered pump, so it stops the moment the grid does. Over a multi-day outage in a warm or brightly lit room, evaporation keeps going while the top-off doesn’t, and salinity creeps upward. Two honest caveats: there’s no single reliable “SG per day” figure to quote, because the rate depends on your tank’s surface area, room temperature, and how low the water already was, so measure rather than assume; and when power returns, top off gradually with fresh RO/DI water, not more saltwater, and re-check with a refractometer before you trust the number.

Battery Air Pumps: What’s Actually Published

Aeration is the piece a battery-powered pump solves automatically, instead of you setting an alarm every 4–8 hours. It’s also why this pump belongs in a permanently staged kit rather than a someday cart: if the outage comes from a no-warning event like an earthquake or tornado, there’s no shopping window at all, and a pump that switches itself on keeps working even if nobody is home when the grid drops. Here’s what the manufacturers actually publish, side by side, including the one product where the spec sheet has a real gap.

Pump	Battery	Runtime (published)	Tank capacity	Source
Aquarium Co-Op Battery Backup (single outlet)	Rechargeable lithium-ion	20 hrs continuous / up to 40 hrs power-save	Not published	aquariumcoop.com
Penn-Plax Silent-Air B11	2x D batteries	72 hrs or more per Penn-Plax (SAB11); up to 48 hrs per the retail listing	Up to 29 gal	Penn-Plax (runtime, batteries, capacity); Amazon listing (the 48-hour figure)
Penn-Plax Air Pod (APB1)	4x D batteries	Up to 150 hrs (retail listing only; not published by Penn-Plax)	Up to 55 gal	Penn-Plax (battery, capacity); Amazon listing (runtime)
Aquatop BREZA AC-DC-ONE	2x D batteries	No published spec	Not published	aquatop.com

The Air Pod’s 150-hour figure is the standout for a genuine multi-day outage, and it needs a caveat we have now been able to make precise rather than vague. We went to Penn-Plax’s own product page for the Cascade Air Pod (APB1). Penn-Plax publishes the battery count (4 D batteries, not included), the capacity (tanks up to 55 gallons) and the power-sensing automatic switchover. It does not publish a runtime figure at all. So the 150 hours is a retailer-listing number, not a manufacturer specification, and this page has corrected itself where it previously called it manufacturer-claimed. Treat it as unconfirmed until Penn-Plax documents it, and if that exact number is what decides your purchase, that is the one spec here you should not lean on. The Aquatop BREZA has the highest published airflow rate (2.5 L/min) of anything reviewed here, but Aquatop doesn’t publish a runtime or capacity figure at all, which is a real information gap, not an oversight on our part.

The B11 needs the opposite correction, and we are making it here because we got it wrong first. This page previously said both Penn-Plax runtimes were retail-listing numbers. They are not. Penn-Plax lists that pump under the SKU SAB11 rather than the “Silent-Air B11” name the listings use, and on that page the manufacturer writes that depending on the batteries used it “can provide your fish with an uninterrupted supply of oxygen for 72 hours or more.” That is a manufacturer figure, and it is longer than the up-to-48-hours the retail listing carries, so the correction runs in the direction of the pump being better documented than we credited, not worse. Two things follow. Provision against 48 rather than 72, because Penn-Plax conditions its number on the cells you put in and publishes no test conditions for it, and stage a fresh pair of D batteries with the pump either way. And treat the wider lesson as the durable one: a name that does not match the manufacturer’s SKU is the single most common reason a spec looks unpublished when it is not, which is why the sourcing note on the next table now names the SKU we checked.

Runtime Per Battery: Comparing These Pumps Fairly

Raw runtime hides a variable: how many batteries it took to get there. A 150-hour pump that eats four D batteries and a 72-hour pump that uses two aren’t as far apart as the headline numbers suggest. The table below normalizes each manufacturer’s published total runtime by battery count, so you can weigh cost-per-outage rather than just the biggest number on the box. One caveat, stated plainly: the per-cell figure is our own arithmetic (published total divided by battery count), a rough comparability aid, not a spec the manufacturer prints, and not a claim that the pump runs on a single cell.

Pump	Batteries	Published total runtime	Runtime ÷ battery count (derived aid)	Notes
Aquarium Co-Op Battery Backup (single outlet)	1 internal rechargeable Li-ion pack	20 hrs continuous / up to 40 hrs power-save	Not on the disposable-cell scale (rechargeable)	Recharges instead of needing fresh cells; cheapest over many outages if you keep it charged
Penn-Plax Silent-Air B11	2x D	72 hrs or more per Penn-Plax (SAB11); up to 48 hrs per the retail listing	~36 hrs per D cell on the manufacturer figure, ~24 on the listing figure	Two-cell simplicity; cheapest to re-provision mid-outage; the only D-cell runtime here the manufacturer publishes
Penn-Plax Air Pod (APB1)	4x D	Up to 150 hrs (retail listing, not a Penn-Plax spec)	~37.5 hrs per D cell	Most total runtime and the best per-cell efficiency of the D-cell pumps, on a runtime figure the manufacturer does not confirm
Aquatop BREZA AC-DC-ONE	2x D	Not published	Not calculable (no runtime spec)	Highest published airflow (2.5 L/min), but no runtime figure to divide

The honest read: the Air Pod (APB1) is the longest-running of the disposable-battery pumps on the numbers as published, and it leads the per-cell column at roughly 37.5 hours, but the gap is now nearly nothing. The B11 sits at roughly 36 hours per cell on Penn-Plax’s own 72-hour figure, and that figure is the one with a manufacturer behind it while the Air Pod’s 150 hours is not. Read those two rows as a tie decided by sourcing rather than by arithmetic, and the B11 wins that tiebreak. The Aquarium Co-Op unit trades disposable-cell math for a rechargeable pack that wins over many outages if you keep it topped up, and the Aquatop can’t be scored on runtime at all, because that number simply isn’t published. One sourcing note on the whole table: the Aquarium Co-Op, Aquatop and B11 runtimes come from those manufacturers’ own pages (the B11 under Penn-Plax’s SKU SAB11), the Air Pod’s 150 hours comes from a retail listing rather than from Penn-Plax, the B11’s competing 48-hour figure is a retail-listing number too, and the per-cell column is our arithmetic on top of all of them.

Reptiles: Why There’s No Single “Safe Hours” Number

Search for “how long can a reptile go without heat” and you’ll find confident-sounding numbers everywhere. Here’s the honest version: no single vet-school or peer-reviewed source publishes one universal figure for reptiles as a group, because tolerance varies enormously by species, size, and room temperature. What does exist is a handful of species-specific figures, at different levels of authority, and the page owes you the difference.

• Bearded dragons: no vet-school, university herpetology, or veterinary-reference source we checked (including VCA and Merck) publishes a specific hours-without-heat figure for this species. A 24-hour number circulates on pet-content sites, but it traces only to other content sites, not to a clinical citation, so this page won’t repeat an unsourced survival number. What is documented: normal husbandry keeps the cool zone at 75–80°F and the basking zone at 90–100°F, per VCA Animal Hospitals. Treat a room sliding toward that cool-zone floor, plus any hypothermia sign below, as the point to act.
• Snakes: roughly 6–12 hours is commonly cited by reptile-keeper sources, provided ambient room temperature stays above about 55°F, with cold-sensitive species (some boas and pythons) faring worse and cold-tolerant species (corn snakes, some skinks) faring better. This range comes from breeder and hobbyist consensus, not a vet-school or peer-reviewed citation, and should be read that way.
• Reptiles broadly, per an avian/exotics vet: brumation-level cold can be tolerated for “a few days” short-term, but prolonged cold raises real illness risk, per Dr. Laurie Hess, DVM, a board-certified avian specialist.

Don’t average these into one number, and don’t read the missing bearded-dragon figure as license to wait longer: the honest gap there is about missing data, not about the species being hardier. A snake’s 6–12 hour figure is keeper consensus, not a clinical citation either, and none of these numbers, present or absent, are an excuse to delay action if your specific animal is showing symptoms.

Minimum Temperature and Hours Without Heat, by Species

Because those tolerances live in different sentences (and at different levels of authority), here they are side by side, with the source tier attached to each so you can weigh them honestly. The one clinical, cross-species rule, a body temperature approaching freezing or any hypothermia symptom is a vet emergency, is in the next section and overrides every row below.

Species	Normal husbandry temp	Emergency room-temp floor	Rough tolerance without heat	Source (authority tier)
Bearded dragon	Cool zone 75–80°F, basking 90–100°F	Not published by any source we checked	No published figure; watch the habitat temperature and the animal, not a clock	VCA Animal Hospitals (husbandry range only; VCA publishes no emergency-tolerance figure)
Ball python	Basking 88–92°F, warm side 80–85°F, cool side 76–80°F	Ambient above ~55°F	~6–12 hrs; sustained below ~72°F risks digestive shutdown	ExoPetGuides / Regarding Reptiles (care-guide and keeper consensus)
Leopard gecko	Warm/belly 88–92°F, cool side 73–76°F, night floor ~68°F	Room above ~60°F	A few days if the room stays above 60°F; first ~12 hrs uneventful	Reptile Craze (care-guide consensus)
Other snakes (general)	Species-dependent; mid-70s°F typical at night	Ambient above ~55°F	~6–12 hrs; cold-sensitive boas and pythons fare worse	Reptile-keeper consensus

Read the tiers, not just the numbers. None of the four rows trace to a vet-school or clinical duration figure: the bearded dragon’s husbandry range comes from VCA, but VCA, like every other primary source we checked, publishes no emergency-tolerance number to pair with it, and the ball python, leopard gecko, and general-snake figures come from care guides and keeper consensus, which is real-world experience, not clinical data. Match the row to your actual species and room temperature, watch for symptoms rather than a clock, and read the missing bearded-dragon figure as this page being honest about a real gap, not an oversight.

The Vet-Emergency Threshold

This is the one number in this section with a clear clinical source, and it matters more than any of the species estimates above: a reptile body temperature approaching 32°F (freezing), or any reptile showing the hypothermia signs PetMD lists, is a medical emergency. PetMD’s wording on the temperature is this: “The closer the reptile’s body temperature is to freezing (32 degrees F) the more of a medical emergency it will be.” The signs it publishes are lethargy, decreased activity, decreased appetite, an increased amount of sleeping, poor growth rate, decreased amount and frequency of defecation or urination, secondary infections, and death. Contact your veterinarian immediately, per PetMD (Sean Perry, DVM). Chronic cold exposure also suppresses reptile immune function, which PetMD says can lead to secondary infections that are life-threatening even after the hypothermia is corrected. So “he warmed back up and seems fine” isn’t the end of the story if the exposure was prolonged.

One honest note on that list, because this page’s promise is that a citation covers exactly what it appears to cover: those are PetMD’s items, in PetMD’s words. An animal that is unresponsive or that you cannot rouse is obviously an emergency too, but that is common sense rather than something PetMD publishes in this list, so we are not going to smuggle it inside the citation.

Never attempt rapid re-warming. PetMD is specific on this: warming a hypothermic reptile too fast can cause life-threatening complications. Gradual, temperature-controlled rewarming, and a vet call, is the correct response, not a heat lamp cranked to maximum.

Keeping Reptiles Warm Without Grid Power

• Insulate first. Extra towels, blankets, or moving a smaller enclosure to an interior room (away from exterior walls and drafts) costs nothing and buys real time.
• Heat packs, with a barrier. UniHeat’s manufacturer-published durations (20 to 120 hours) give you a genuine multi-day option, and UniHeat’s own published temperature chart for the 40-hour pack peaks just under its 150°F gridline, so always place a towel or sock between the pack and the animal or enclosure surface. UniHeat’s own FAQ points the same way: “DO NOT place shipping warmer directly on or beside sensitive living things, as they may be harmed.” Every published figure for these packs, including where the publishers contradict each other about the same pack, is on our heat pack runtime page.
• SnuggleSafe pads, once you can charge one. The manufacturer rates it for up to 10 hours of warmth per ~5-minute microwave charge. The catch: you need a working microwave, so this is more useful once a generator is running or grid power partially returns than in the first hour of a blackout. SnuggleSafe markets this for dogs, cats, and small mammals; reptile use is anecdotal, not manufacturer-endorsed, so treat it as a bridge option, not a primary plan.
• Never use a standard human hand warmer as a reptile heat source. The prohibition is unconditional, and there are two separate reasons for it. The first is burn risk, and it has a manufacturer number behind it: HotHands, a widely available air-activated hand warmer brand, publishes an average temperature of 135°F and a maximum of 158°F, for up to 10 hours. That is a product engineered against a human palm through a glove, not against reptile skin. The second is oxygen: these warmers heat by iron oxidation, so they consume oxygen from the air around them, which is a problem in a closed enclosure specifically, per reptile-keeper sources. An unsealed enclosure removes the second reason and leaves the first one completely intact, so “not in a sealed container” is not the rule. “Not against the animal” is.
• Keep exotic pets away from combustion fumes. If you’re improvising warmth elsewhere in the house with candles or a propane heater, keep the reptile or aquarium room clear of those fumes entirely, per Dr. Laurie Hess, DVM.
• ASPCA’s kit guidance: include a sized soaking bowl and a heating pad or other warming device in your reptile’s kit ahead of time, and fill bathtubs or sinks before an outage hits for backup water access.

Hedgehogs fall in the same temperature-dependent category as reptiles here, since a blackout takes away the heat source they rely on. Our hedgehog emergency preparedness kit covers the no-power backup heat source and the temperature line that means it is time to act.

When Cold Is Not the Failure Mode

Everything above this line assumes the animal is losing heat. For part of this category that assumption is backwards, and following the warming advice on this page would be the injury rather than the fix. Each case below has its own tell.

Cool-water species. The University of Kentucky’s Ambystoma Genetic Stock Center, which maintains the reference axolotl colony, holds its animals at 15 to 17 degrees Celsius, roughly 59 to 63°F. Its own husbandry guide says it does that using the building HVAC system plus two 6-ton auxiliary air-cooling units. Read that as a statement about what the emergency is: the institution that keeps the species spends cooling capacity to prevent it, so for a cool-water animal the outage that matters is the summer one that stops a chiller or the air conditioning, not the winter one that stops a heater. The Merck Veterinary Manual makes the same point in general husbandry terms for amphibians, that water chillers and air conditioning have to be considered for the enclosure rather than treated as optional. We are not going to publish an upper-lethal temperature for these species, because we have not found a clinical figure to cite for one. Check the animal’s own husbandry source for its range, and treat every warming instruction on this page as not applying to it.

Humidity, not just temperature. An outage takes out misters, foggers, ultrasonic humidifiers and heated water at the same instant it takes out the heat, so a screen-top enclosure starts drying at the same time it starts cooling. Merck treats this as a first-order variable rather than a refinement: “Humidity that is too high or too low can create serious problems.” Inadequate humidity contributes to dysecdysis, the shedding failure, and to shell pyramiding in chelonians. For amphibians Merck is blunter still, that they need moisture to prevent desiccation. The practical overlap is convenient. Covering an enclosure to hold heat holds humidity too, so the blanket is doing two jobs at once, and the shallow water dish that was optional yesterday is not optional now.

The animal’s own thermal direction. The species tolerance table earlier on this page runs one way, toward cold, because that is the direction the animals in it fail. Before you apply any of it, confirm your species belongs on that side of the line. A keeper who has only ever read winter-outage advice will reach for a heat pack in a July blackout for an animal whose entire husbandry is built around keeping it cool, and that is a reachable mistake, not a theoretical one.

Battery Pump, Power Station, or Move the Animal

Three options cover nearly every outage, and owners tend to compare them on price when the thing that actually separates them is which failure each one can carry. An air pump cannot make heat. A power station can, for a while. Moving the animal solves both and costs you the setup.

Option	The failure it carries	How long, per published specs	What it cannot do
Battery backup air pump	Oxygen, and only oxygen	20 to 72 hours across the pumps compared above, with one retailer-listing-claimed 150-hour outlier that Penn-Plax itself does not publish	Anything thermal. A sub-2-watt pump is not in the same class of load as a heater, and no amount of runtime changes that.
Portable power station or generator	Heat, filtration and oxygen together	Hours rather than days once a 100 to 200 watt heater is on it, at Aqueon’s 5 watts per gallon	Run indefinitely, or run indoors. A generator brings the carbon monoxide rule below along with it, and that rule is not negotiable.
Move the animal or the water	Both, permanently	NC State’s transport guidance: sparsely packed fish, about five inches of fish per gallon, in a bag of roughly one-third water and two-thirds air, should survive at least 36 hours, and NC State conditions that figure on the external temperature staying adequate	Be undone in an afternoon. Breaking a tank down is a one-way decision for that day.

The standing costs differ in kind rather than in degree, which is exactly why a head-to-head on price misleads. A pump costs you either fresh D cells in a drawer or the discipline to keep a rechargeable pack topped up. A power station costs you the recharge cycle, and a generator costs you fuel plus the twenty-foot rule every single time. Moving costs nothing to own and a great deal on the day you do it, and NC State’s discard-and-weigh figure in the escalation list above is the closest thing to a cost you can actually compute for it before you commit. If you are choosing backup power for the whole household rather than for one tank, our quiet generator versus portable power station comparison is the fuller version of that decision.

The rule that falls out of the table: pick by the load that failed, not by the gear you like. If the water is holding temperature and it is the pumps that stopped, a battery air pump is the entire answer and a power station is overkill. If the room is dropping and the heater is what matters, no air pump helps at all, and the real choice is between backup power and leaving.

When to Relocate the Animal

Insulation and heat packs buy time; they don’t fix a house that’s genuinely losing heat for days. Move fish or reptiles to a warmer location (a neighbor’s powered home, a pet-friendly shelter with exotic-animal capacity, or a running vehicle as a short bridge, under the rules in the next section) when:

• Your heat packs or battery pump are running low and you have no way to replace or recharge them.
• A reptile is showing any hypothermia symptom: don’t wait for any duration figure you may have seen elsewhere, including for species this page won’t put a number on; symptoms override the clock.
• The outage has a real end-date beyond your gear’s published runtime (72 hours of UniHeat plus no clear restoration timeline, for example).
• Room temperature itself is dropping toward the low-50s°F or below for reptiles, or toward the high 50s°F for fish, with no sign of stabilizing.

If the Warm Place Is a Running Vehicle, Read This First

A car is the warm space most people reach for at 2am in a winter outage, and it is the only one on that list that can kill the person carrying the animal to it. It is a legitimate option. It is not a free one, and the rules below are not optional garnish on it.

Never run the engine inside a garage, and an open garage door does not fix it. The EPA states it directly: “DON’T idle the car in a garage” even if the garage door to the outside is open, because “fumes can build up very quickly in the garage and living area of your home.” That is the same discipline this page already applies to generators lower down, from the same reasoning: a garage is indoors with a door, not outdoors with a roof. Back the vehicle fully out before it runs.

Clear the tailpipe before every start, and crack a window while it runs. If the car has been sitting through a snowstorm, the National Weather Service’s stranded-driver procedure is the one that applies: “Clear snow from the exhaust pipe to avoid gas poisoning.” And while the motor is running, NWS says to “open the window a little for fresh air to avoid carbon monoxide poisoning.”

A low enclosure is not a safer layer of air. The ASPCA lists vehicle exhaust among the common carbon monoxide exposure sources for pets, and its prevention list carries the rule “Never run engines in a closed area”. It also notes that carbon monoxide is roughly the same density as room air, so animals close to the ground are at no more risk than taller ones, which cuts both ways: a tank on the floorboard or a carrier in the footwell is not in a protected pocket either. Birds are the exception the ASPCA singles out as more sensitive still.

Treat it as a bridge, never as a place to leave the animal. An unattended vehicle swings from too cold to too hot with nobody watching the thermometer and nobody there to catch a symptom. If the destination is a neighbor’s powered house or a shelter, the car is the trip, not the plan. Moving a tank rather than an animal is a transport job with its own numbers, which are in the bag-and-bucket section above.

Our blizzard stranded-car guide carries the full engine-cycling protocol for a vehicle you cannot leave, and our generator carbon monoxide guide covers the identical mechanics when the engine is a generator instead of a car.

After the Power Comes Back: Confirm Every Load Actually Restarted

This is the failure nobody drills for. The grid returns, the lights come on, the refrigerator hums, and the under-tank heater, the ceramic emitter or the aquarium heater stays dark, because the switch it was plugged into came back in the off position. Every other page about outages ends when the power returns. For this category that is the moment a second, quieter emergency starts, and it runs while you believe the problem is solved.

It is usually not a defect. On most smart plugs it is a documented setting, and the vocabulary differs by brand, which is a large part of why owners never go looking for it.

Brand line	What the setting is called	What the manufacturer documents
SONOFF, through eWeLink	Power-on state	ON, OFF, or KEEP, where KEEP means the state the device was in before the power was cut. ITEAD’s own support article names the default as OFF.
TP-Link Tapo	Default State	Last On State, On, or Off, set per device under Device Settings, per TP-Link’s own announcement of the feature. That announcement also carries a prerequisite worth acting on before an outage rather than during one: “Please ensure your Tapo Plug and Tapo app is up to date to use this feature.”
Meross	No equivalent setting named in the public FAQ	The Meross FAQ answers a narrower question than owners read into it: devices retain their settings and reconnect to the network once power is restored. Retaining a setting and energizing a socket are not the same claim.

None of what follows requires you to know which brand you own.

Set it to ON, not to the last-state option. Last state is only the right answer if the load happened to be energized at the instant the grid dropped. A heat source on a thermostat is cycling by design, so it is off for part of every cycle, which makes last state a coin flip on precisely the load you cannot afford to lose. For anything an animal’s body temperature depends on, the honest setting is unconditional ON, with the actual temperature control left to the thermostat where it belongs.

Then test it, because until you have watched it work the setting is a claim. Pull the plug at the wall, wait a minute, plug it back in, and watch whether the load energizes on its own without you opening an app. Repeat after any firmware update, since this behavior lives in firmware rather than in the app. And if you want the honest version of the test, run it with your router switched off as well: a plug that needs the network to decide what to do is a plug that needs your router and your internet service to be back first, and in a real outage those return on their own schedule rather than yours.

For a heat load, prefer a controller with nothing to negotiate. A plug with a physical rocker switch tells you by sight and by feel that the contact is closed, with no stored state to restore. A time switch with a battery-backed clock is the same idea one level up: Intermatic sells its electronic 24-hour time switches with battery carryover and a battery-powered clock rated at three years minimum, so the schedule survives the interruption instead of restarting from nothing, and the same units carry a MANUAL position that overrides the program outright. The trade is worth naming rather than glossing: carryover is sold as a differentiator, which is itself a statement about what the units lacking it do, so a plain mechanical dial with no battery carryover has nothing keeping its clock running during the outage and needs resetting even though the load itself came straight back.

None of this replaces the thermostat. Merck’s reptile husbandry guidance puts a thermostat between the heater and the animal for good reason, screened from the animal and positioned toward one end of the enclosure so there is still a thermal gradient. A smart plug is a switch, not a temperature control, and treating one as the other is how a heat source ends up either dark or unregulated. The only question this section settles is whether the thermostat gets its supply back.

The tank has its own restoration rules, and the two that are time-critical are already above rather than repeated here: the feeding hold in the staged timeline table, and the fresh-water top-off rule in the salinity section. Bringing filtration, biological media and the nitrogen cycle back up has an order to it, and that order is a longer sequence than an outage playbook should try to carry in passing.

Generator Safety: The One Mistake That Kills People, Not Just Fish

If you’re running a generator to keep an aquarium heater or reptile lamp powered, the carbon monoxide risk applies to your whole household, not just the animals. The Consumer Product Safety Commission is direct: never run a portable generator indoors, in a garage, basement, crawlspace, or shed. Run it outside, at least 20 feet from the house, with exhaust directed away from windows, doors, and vents, and look for a model with a built-in CO shut-off feature. That single mistake has killed far more people than any outage-related pet emergency.

The size of it, stated the way the agency states it: according to the U.S. Consumer Product Safety Commission, an average of nearly 100 people die in the U.S. each year from carbon monoxide poisoning caused by portable generators. Three qualifiers travel with that number and are worth carrying, because a bare “100” reads more precise than it is. It is a 2019 to 2021 annual average, not a single year’s body count, published in CPSC’s Non-Fire Carbon Monoxide Deaths Associated with the Use of Consumer Products, 2021 Annual Estimates (Directorate for Epidemiology, February 2025). It is a statistical estimate scaled to death-certificate data rather than a raw incident tally, and CPSC notes recent years are preliminary and typically revised upward. And its scope is consumer portable generators only: it excludes fire-caused CO, motor-vehicle exhaust, work-related incidents, and the built-in generators on RVs and boats. If you see an older figure in the 80s quoted elsewhere, that is not an error someone invented, it is a superseded CPSC estimate from an earlier window and a different counting method. The trend across the revisions runs upward.

Building an Aquarium/Reptile Outage Kit

• A battery backup air pump, sized to your tank (see the comparison table above): this is the single most useful item for fish.
• 1–2 chemical heat packs per animal or per 20–30 gallons of water, plus a towel or sock to use as a barrier. That count is retailer guidance rather than a tested figure, so check the tank or enclosure temperature instead of trusting it, and note that UniHeat’s own 40-hour pack page tells shipping customers to use only one pack per shipment to avoid overheating. It is a stocking figure for a drawer, not a dose.
• A battery-powered or manual thermometer you can check without opening the enclosure repeatedly. This is also the instrument the survival timer above runs on, and that timer’s whole input is a small difference between two readings an hour apart, so you want one that reads in the dark and resolves to a tenth of a degree. Those are two requirements, not a trade: a display you can’t read at hour six is useless, and so is one that rounds away the half-degree per hour you are trying to measure.
• A net and a heavy-duty bag, kept together. NC State’s supply list carries both as separate line items, and its note on the net is what explains why they are used together: herding a fish into the bag with the net is safer than lifting with the net alone, and NC State’s stated reason is damage to the protective mucus layer and epidermis.
• A soaking bowl sized for your reptile, per ASPCA guidance, plus a backup water source (fill the tub ahead of a forecasted outage).
• A printed copy of your species’ normal temperature range, so you’re not searching for it on a dead phone at hour six.
• Your vet and, where relevant, an exotics-specialty vet’s phone number, written down, not just saved in a phone that might be dead.
• A hygrometer, if humidity is part of your animal’s husbandry, since the same outage that stops the heat stops the misting.
• A tested power-restore setting on every plug in the chain, plus, for a heat load, a switch you can operate by hand. This is a preparation task rather than a purchase: the test in the restoration section above takes two minutes on a normal Tuesday and cannot be run during the outage it protects you from.

For the broader household picture (what to do the moment the power goes out for every animal in the house, not just tanks and terrariums), see our pets and power outages pillar guide, our breakdown of generator carbon monoxide risks for pets, and portable power stations for pets if you’re evaluating whole-system backup power rather than single-device solutions. If medication refrigeration is also part of your outage plan, see pet medication and refrigeration during an outage.

Tanks aren’t the only plugged-in thing that stops. An automatic feeder and a circulating water fountain go dark in the same second the pump does, and their battery backup, where it exists at all, rarely lasts as long as owners assume. Our guide to how long an auto feeder lasts in a power outage covers which units keep running, how to test yours before you need to, and the no-power fallback.

How We Chose

This page is spec-and-evidence analysis: manufacturer product pages and support documentation, a veterinary college’s public guidance, a clinical veterinary reference, a university research colony’s published husbandry guide, vet-reviewed consumer sources, and a federal safety agency, not hands-on product testing. The power-restore behavior described in the restoration section comes from each vendor’s own support material rather than from testing a plug of every brand, and where a vendor does not document the setting we say that instead of inferring it. We don’t sell aquarium or reptile equipment or take sponsorships; we’re reader-supported through the disclosed Amazon links above, and commissions never touch which products we recommend. Every number above is either sourced to a named page or explicitly flagged as unpublished, community-sourced, or unverifiable in this research pass. Full methodology at /review-methodology.