The Electric Well Pump Stops Instantly: The Numbers That Decide What Your Animals Drink

The power goes out at nine at night. Somewhere in the house there are three bowls, a cat fountain, a fifty-five gallon tank with a filter in it, and a dog that drinks half a bowl every time she comes in from outside. On a municipal system, none of that is urgent. The tap still runs.

On a well, the tap stopped when the lights did. There was no warning sound, no sputter, nothing to notice. The next person who turns the kitchen faucet gets a few seconds of water at falling pressure and then air.

This page is about the four or five decisions that follow, and it is written around three quantities that decide all of them. How much water is genuinely left in the system. Whether you should use it. And what it would take, electrically, to start the pump again. Each of those three has a published answer from a manufacturer or an agency, and the answers are more specific and more useful than the general advice suggests.

What This Page Adds, and What This Site Already Said

We should be straight about the starting point, because this site is not silent on well pumps and never has been.

Our pets and power outages hub already carries a dedicated section on well water, a standalone FAQ answer about a well pump with no power, a row in its dependency table naming the electric well pump as one of four things that genuinely require backup power, a checklist line about storing water if your home is on a well, and a closing paragraph that names the well pump among the four dependencies a household should identify in advance. Its position is that the fix is stored water rather than improvisation, and that a generator or solar supply tied specifically to the pump is the documented answer for extended outages. Our inland hurricane page goes further and publishes the distinction that matters most here, telling readers to plan for the draw-down rather than for the tank. Our generator versus power station comparison names a well pump as one of the loads where a fuel generator earns its place rather than a battery.

All of that is correct and none of it is what this page is for. What none of those pages does, and what nobody had done here, is publish the numbers. How many gallons is a draw-down. How many watts is a well pump. What, specifically, would have to be true of a machine for it to start one.

This page publishes those three, from the tank manufacturer, the motor manufacturer, and the agencies that regulate private wells. It reaches one conclusion the other pages could not reach without them, and it is about voltage rather than capacity. And it inherits one instruction from a state environmental agency that changes what the gallons figure even means, which is the part this page most wants you to read.

Two boundaries, stated up front so you can leave early if you are on a different problem. This page does not size your animals’ water supply; that arithmetic, and the argument about which per-pound rate to use, lives on how much water per dog in an emergency. And this page is about a well whose water is fine and whose pump has no electricity. A well that has been flooded or otherwise contaminated is a different hazard with a different clock, and the instructions below do not cover it. For that case, our pet water after a flooded private well guide covers what changes when the water itself, not just the power, is the problem.

The Pump Stops Instantly, and the First Action Is a Breaker

There is no buffer between the outage and the pump. PrivateWellClass.org, which describes itself as a collaboration between the Rural Community Assistance Partnership and the University of Illinois through the Illinois State Water Survey and the Illinois Water Resource Center, puts it in one sentence: “If you have an electrical well pump, you cannot pump water from your well during a power outage.”

The Ohio Department of Health, on its private water systems page dated August 28, 2018, describes the same moment from the system’s side: “During a power outage, the pumping systems will not operate and a limited quantity of water is stored in the pressure tank and water lines.” Ohio also notes that private water systems in that state are not all wells, listing springs, rainwater cisterns and ponds alongside them, and that changes the answer below for some households.

The first action, though, is not about water at all. Both PrivateWellClass and NEHA’s Private Water Network tell owners to cut power to the pump, for two different reasons.

PrivateWellClass: “It is important to turn off the power supply to the pump, so it does not start automatically when the power is restored.” Its companion tip is the kind of thing that only ever gets done on a calm afternoon: “Clearly label your circuit box, so you quickly know which breaker turns off your pump.”

NEHA’s Private Water Network factsheet, produced under a cooperative agreement with the CDC, gives the safety version, instructing owners to make sure they turn off power to the well pump before inspecting the area around the well because of the danger of electrical shock and damage to the drinking water system. Its Before a Power Outage list carries the same step, adding the detail that the switch may be by the water pump or tank, or at the main electrical panel.

For a household with animals this matters for an unglamorous reason. The moment the power comes back, an unattended pump restarts into a system that may have been sitting depressurized for hours, and the person who would normally notice is asleep or at work. Cutting the breaker makes the restart a decision rather than an event.

Everything about breakers, transfer switches and generator connections in this article is a description of what manufacturers and agencies publish, not an instruction to perform electrical work. That work belongs to a licensed electrician or a licensed well contractor.

What the Tank Actually Holds: Drawdown Is Not Tank Volume

Here is the single most useful thing on this page, and it is a number most well owners have never seen even though their own tank’s manufacturer publishes it.

A pressure tank is not a water tank. Penn State Extension, in a publication by Bryan Swistock, its Water Resources Coordinator, says so plainly: “The pressure tank allows a water system to operate automatically. It is, in a sense, a storage tank–but it has very limited storage capacity. Its primary purpose is to create and maintain pressure on the water in the pipeline.” The double hyphen in that sentence is Penn State’s own punctuation, reproduced here as it appears.

PrivateWellClass explains the mechanism in a companion article dated February 19, 2026: “A pressure tank serves two basic purposes: First, it provides water storage and water under pressure to your home when the pump is not running.” On the cycle, it states: “When the pressure in the line drops below the low-end preset level, generally 20 to 40 PSI, it engages automatic controls and preset gauges like the pressure switch in Figure 1.” The pump then runs, in that article’s own words, “until the usage stops and or the pressure in the line reaches the preset high-end level, generally 40 to 60 PSI.” The phrase “and or” is the source’s own wording and we have not tidied it.

That cycle is the whole reason a tank’s volume and its usable water are different numbers. The water that leaves the tank between the high pressure and the low pressure is what comes out at working pressure with the pump off. That quantity has a name, drawdown, and it is a published specification.

The current numbers, by named model

Amtrol makes the Well-X-Trol line of residential well tanks. Its specification card, form MC7025 (03/24), carries copyright 2024 Worthington Enterprises and is served from amtrol.com. Tank volume and drawdown are separate columns on the same table, and drawdown is given at three pressure-switch settings.

Model (Amtrol Well-X-Trol)	Tank volume, gallons	Drawdown at 30/50 psig	Drawdown at 40/60 psig	Drawdown at 50/70 psig
WX-201	14.0	4.3	3.8	3.3
WX-202	20.0	6.2	5.4	4.7
WX-202XL	26.0	8.0	7.0	6.1
WX-203	32.0	9.9	8.6	7.6
WX-250	44.0	13.6	11.8	10.4
WX-251	62.0	19.2	16.6	14.6
WX-350	119.0	36.8	31.9	28.1

Every figure above is Amtrol’s, read off form MC7025 (03/24) on August 15, 2026. The same sheet carries the manufacturer’s own caveat, and it belongs with the numbers wherever they travel: “Drawdown can be affected by various ambient and system conditions, including temperature and pressure.” It also gives the factory precharge as 38 psig and the maximum working pressure for all models as 150 psig.

Three observations from those rows, and all three are this site’s arithmetic on Amtrol’s published table rather than statements Amtrol makes.

First, drawdown runs close to 27 percent of tank volume at a 40/60 switch across the whole residential range, from the 14 gallon WX-201 up to the 119 gallon WX-350. That consistency is useful, because it means a household that knows only its tank’s nominal size can estimate within a few gallons.

Second, the switch setting moves the answer by roughly a quarter. The WX-250 gives 13.6 gallons at 30/50 and 10.4 at 50/70. Higher house pressure buys less reserve. If you do not know your setting, the gauge near the tank shows it, and it is worth writing down.

Third, the difference between the tank on your basement wall and the water in it is large enough to change a plan. A person who believes they have a 44 gallon reserve and actually has 11.8 has made a four-fold error.

A second figure, from an extension service, that does not match

Penn State Extension publishes a different number, and this page is not going to reconcile them, because they describe different things and averaging them would produce a figure neither source stands behind.

Penn State: “About 20 percent of the capacity of the pressure tank contains usable water. A 42-gallon tank will discharge about 8 gallons before the pump starts; (an 82 gallon tank–16 gallons; a 120 gallon tank– about 24 gallons).” The punctuation, including the uneven spacing around the double hyphens, is Penn State’s own.

The sentence Penn State puts immediately after that one is worth carrying with the figure, because it is the conclusion Penn State draws from it: “Thus, larger pressure tanks alone will provide slightly larger amounts of stored water, but the increased storage is not enough to solve problems with a low-yielding well.” Buying a bigger tank is not a water-supply strategy, in that publication’s judgement, and it is not one here either.

That figure is roughly 19 percent, against Amtrol’s roughly 27 percent at 40/60. Here is what we can see that might explain the gap, offered as this site’s reading of the two documents rather than as a claim either makes. The Penn State passage immediately preceding that figure describes a tank in which “the air in the space above the water is compressed” directly, with the switch stopping the pump at about 40 psi and restarting it at about 20 psi. That is a plain air-over-water tank on a 20/40 switch. Amtrol’s table is for diaphragm tanks with a sealed air charge, at 30/50 and above. PrivateWellClass’s tank article describes exactly that generational difference, noting that in the original design “the air can dissolve into the water, lowering the air pressure in the tank and reducing the amount of air in the tank” and that most new tanks are bladder tanks with a permanent separation.

The practical consequence for a household is simple and does not require resolving the disagreement. Look up your own tank’s model number rather than applying either percentage, and if your tank is old enough to be an air-over-water design, expect the lower end.

One more note on document currency, because it is a check this site runs and it produced a real result here. We compared form MC7025 (03/24) against the prior edition of the same form, MC 7025 (04/14). The drawdown figures for every model named in the table above are identical across the ten years between them. Two entries did change: the WX-103’s tank volume is listed as 6.7 gallons on the current sheet against 7.6 on the older one, with its drawdown falling accordingly, and the WX-252 appears on the older sheet and not the current one. Cite the current form, and check your own model against it rather than against a figure copied from an older document.

Why we are not converting either figure into days of pet drinking water

This is the point where a page like this normally does the obvious arithmetic. Take the drawdown, take a per-pound daily water rate for a dog, divide, and publish a number of days.

We are not going to, and the reason is the whole next section. Michigan’s Department of Environment, Great Lakes, and Energy tells a well owner not to withdraw that water during an extended outage, and the Chester County, Pennsylvania Health Department publishes the same instruction in closely tracking wording, so treat that as one instruction carried by two jurisdictions rather than two independent findings. What follows here rests on Michigan’s instruction alone. Publishing a ration out of a reserve an agency says not to open would be telling you to do the thing the agency told you not to, and telling you to do it to an animal that cannot report the consequence.

What the drawdown figure is genuinely for is different and still valuable. It tells you how fast normal household use will empty the system, which tells you how much time you have before the decision below is forced on you. Penn State’s own Table 1 gives the household side of that arithmetic: a standard toilet at 5 gallons per flush, a low-flush toilet at 1.6, and standard faucets and showerheads at 3 gallons per minute. Set against Amtrol’s WX-202 drawdown of 5.4 gallons at a 40/60 switch, one flush of a standard toilet is nearly the entire reserve and one flush of a low-flush toilet is roughly 30 percent of it. A five-minute shower at Penn State’s standard rate is about three times the whole thing. That arithmetic is ours, joining two published tables that were not written to be read together.

The animals are not what empties the tank. An ordinary evening of a household behaving normally is, and it happens before anyone has thought about the bowls.

The Paragraph This Page Exists For

Michigan’s Department of Environment, Great Lakes, and Energy publishes a two-page fact sheet titled Protecting Your Private Water Supply in an Emergency, dated (04/2020). Its power-outage section is short, and it is the most consequential thing we found for a pet household on a well.

It begins with the mechanism most owners have never been told:

“Under normal operation an electric water pump draws groundwater from the well and maintains the water pressure within the storage tank and piping. Pressurization forms a physical barrier that protects the drinking water against the entrance of contaminants. During a power outage a water well system will lose pressure if faucets are opened. As a result, bacteria, viruses, and other disease-causing organisms can find their way into the water system. The water can become contaminated without any noticeable change in taste, odor, or clarity.”

Read the last sentence twice. The failure this describes is invisible. There is no cloudiness to check for, no smell to catch, nothing that would make a careful owner hesitate before filling a bowl.

Then the instruction:

“During extended power outages it is important to refrain from opening faucets, taking showers, or flushing toilets. While your water system may have several gallons of water in storage, withdrawing water while the power is out depressurizes the system and increases the risk of contamination.”

Notice that EGLE knows about the reserve. It names it, in the same sentence in which it tells you not to use it. The gallons are real and the agency’s position is that spending them is what creates the hazard.

We verified those paragraphs twice, by two routes, on two different hosts, because they carry more weight on this page than anything else on it. The current fact sheet is on michigan.gov and is dated (04/2020). The earlier edition of the same document, published by the Michigan Department of Environmental Quality and marked (Rev. 5/2015), is hosted by the Michigan Association for Local Public Health, and its power-outage paragraphs are word for word identical. The agency’s name changed and the sentences did not.

What happens if the system loses pressure completely

EGLE escalates from advice to a testing requirement when the system runs dry, and it does so twice in the same document.

The softer version: “If your water system loses all of its pressure (no water comes out of the faucet), it is a good idea to have the water checked for coliform bacteria.”

The firmer version, from its temporary water supply section: “If your water system was depressurized during interruption of electrical power or if flooding of the well has occurred, the water should not be consumed until testing has verified that it is potable (fit to drink).”

And the clearance standard, which is worth knowing because it sets the real duration of the problem: “The water is considered safe for drinking after two consecutive samples, collected 8 hours apart, show that coliform bacteria are not present.”

That is a minimum of eight hours between samples, plus laboratory turnaround on both, plus the time it takes to get sample bottles. A household that draws its tank dry on the first night of an outage is not back to normal when the power returns.

The Chester County, Pennsylvania Health Department publishes closely matching guidance for its own residents, with wording that tracks Michigan’s on the mechanism and then diverges usefully at the end. It states that “If your water system loses all of its pressure and no water comes out of the faucets, it is important to check the water quality before continuing to consume the water. It may also be advisable to disinfect your well to ensure that your family has a safe drinking water supply.” And it adds a step Michigan does not: “Before drinking the water, flush all lines letting the water run for a few minutes.” The sentence after that one connects to Ohio’s point below: “If there are concerns that the treatment system is not functioning properly, water should be tested before drinking it.”

We are naming Chester County as a second jurisdiction publishing the same instruction, not as an independent derivation of it. Its power-outage paragraph closely tracks Michigan’s wording, and a county health department relaying a state fact sheet is a relay rather than a separate authority.

The treatment system is off too

Ohio adds a hazard the other documents do not, and it applies to a large number of well households.

The Ohio Department of Health states: “Treatment units also need electricity and will not perform the function of treating the water, making the water unsafe for drinking.” That is the sentence for anyone whose well runs through an ultraviolet disinfection unit, a chlorine injection system, or any other powered treatment stage. On those systems the water reaching the tap during an outage has not been treated, independently of anything happening to the pressure.

Ohio’s page names springs, rainwater cisterns and ponds as other private water sources in that state, and its sentence covers them too. If your household’s water is treated on the way in rather than clean at the source, an outage removes the treatment before it removes the pressure.

The join, and it is this site’s own

Here is the part no single document states, and we are labelling it as this site’s reasoning rather than as a published instruction, because it is a conclusion drawn from putting two sources together.

PrivateWellClass’s method for getting water out during an outage is a faucet at the bottom of the tank. Using it works, and it works by deliberately draining the tank. Michigan EGLE’s rule is that a system that has been fully depressurized should not be consumed from until two clean coliform samples come back. Those two facts, held together, mean that the method for obtaining water is also the method for triggering the testing requirement. The last of the water you draw is, by the second document’s logic, the water most likely to be the water the first document did not intend you to drink.

None of the documents we opened for this page publishes that join, and none of them weighs it for an animal. What follows for a pet household, in our view rather than any agency’s: the reserve in the tank is worth understanding so you know how little time you have, and it is not a plan for filling bowls. The plan for filling bowls is water you already stored.

And if you did not store any

That is the reader this page owes the most to, and the honest answer is not an empty bowl.

None of these documents weighs a night of thirst against a bacterial risk, and neither can we. What we will not do is tell you to leave an animal without water while the question stays open. Our page on how long a dog can go without water carries the site’s position on the other side of that trade, and it is worth reading before you decide to wait.

What follows from that, again as this site’s reasoning rather than any agency’s instruction, is that the water you are looking for tonight is water that does not come from the well. Sealed bottled water already in the house. Containers someone filled before the outage. A neighbour, a relative or a store on municipal water, where the tap does not depend on your electricity and keeps running through a power outage, which is the same distinction this page draws about filling a bathtub. Any of those is off the well and outside the depressurization question entirely.

What this page will not hand you is a way to make the well water itself safe. Disinfection is a contractor and health department procedure with its own concentrations and hold times, and the section further down on what happens after the tank is drawn sets out why we are not publishing one here.

Where the Sources Disagree, Openly

Five publishers address a private well during a power outage, and they do not give the same answer to the most basic question a household has. This table gives one row per issuer and says “Not addressed” where an issuer is silent, rather than borrowing an answer from a neighbouring row.

Source (all read 2026-08-15)	May you draw water from the pressure tank during the outage?	Testing afterwards	What to store beforehand	Does it address animals?
Michigan EGLE, Protecting Your Private Water Supply in an Emergency (04/2020)	No, for extended outages: “refrain from opening faucets, taking showers, or flushing toilets”	Coliform check if all pressure is lost; do not consume after depressurization until testing verifies it is potable; two consecutive samples 8 hours apart	Bottled water; “A few 5-gallon pails filled with water and stored in your basement are helpful for toilet flushes and other nonpotable uses”	Not addressed. The phrase warm-blooded animals appears only inside its definition of coliform bacteria
PrivateWellClass.org, University of Illinois and RCAP (article dated March 19, 2026)	Yes: “If you have a pressure tank, you can use the water in it. Your tank should be equipped with a faucet near the bottom”	Not addressed in this article	Follow the CDC’s emergency water supply guidance; consider a generator or solar backup connected to the pump	Not addressed
Ohio Department of Health, Power Outages and Private Water Systems (page dated August 28, 2018)	Not stated either way; notes “a limited quantity of water is stored in the pressure tank and water lines” and that unpowered treatment units make the water “unsafe for drinking”	Not addressed on the page we could open	Not addressed on the page we could open	Not addressed
Chester County (PA) Health Department, Well Water and Power Outages	No, for extended outages; wording closely tracks Michigan’s	Check water quality if all pressure is lost; disinfection may be advisable; flush all lines for a few minutes before drinking; test annually	Not addressed	Not addressed
NEHA Private Water Network, Power Outages (CDC cooperative agreement OT18-1802)	Not stated either way; during an outage, “Use a safe source of water such as bottled water for drinking, cooking, and bathing”	“Have your well water tested at a certified water testing laboratory after a power outage”; use bottled or stored water until lab results confirm safety	“Buy or store safe water, such as bottled water, to use for drinking, cooking, bathing, and flushing toilets for three days”	Not addressed. The word animal appears only in an instruction to direct flushing water away from animal habitats

Two things follow from that table, and they are worth separating.

The disagreement between the first two rows is real and neither party is careless, and there is a clue in each document that they may be answering different questions. The sentence PrivateWellClass puts immediately before its pressure tank option reads: “Depending on the length of the outage, there are some options for obtaining water available to you.” Michigan attaches its prohibition specifically to extended power outages. Both documents, in other words, appear to think duration matters, and neither publishes a threshold that would tell you where an extended outage starts. Reading them as compatible at short durations and incompatible at long ones is this site’s reading, not a reconciliation either publisher offers, and it leaves the reader without the one number that would settle it.

The second thing is the last column, and it is the finding that most concerns a reader of this site. In every document we opened for this page, and that list includes Penn State Extension’s low-yielding wells publication and both PrivateWellClass articles as well as the five rows above, nothing addresses giving the water to a pet. That is a description of a specific search performed on August 15, 2026, not a claim that no such document exists anywhere. If a state veterinary body or an extension service has published on well water for animals during an outage, we did not reach it.

What the Pump Draws, From the Motor Manufacturer

Now the electrical half, which is where this page’s conclusion lives.

Two pages on this site deliberately publish no equipment wattages. Our generator sizing worksheet tells readers to walk to each device and read its own plate, and our watt-hour budget calculator declines to supply a typical figure for anything, on the grounds that a number borrowed from a similar product is wrong in a way that compounds. Both positions stand, and this section is written to be consistent with them rather than in spite of them.

The figures below are not typical values or estimates. They are the motor manufacturer’s own published specifications, per named model prefix, which is exactly the thing those two pages tell you to go and find. Where a figure appears, its table and model prefix appear with it. Nothing here should be applied to a motor whose plate you have not read.

Table 29, by model prefix

Franklin Electric’s AIM Manual, 2023 edition, is the application manual for its submersible motors. Its Table 29 is headed “4-inch / Single-Phase / Encapsulated Motor Specs / 3450 RPM” and gives, for each model, full load amps and watts, maximum load amps and watts, locked rotor amps and a KVA code.

Model prefix (Franklin AIM 2023, Table 29)	HP	Volts	Type	Full load	Maximum load	Locked rotor amps	KVA code
244504	1/2	115	4“ 2-wire	10.0 A, 670 W	12.0 A, 960 W	64.4	R
244505	1/2	230	4“ 2-wire	5.0 A, 670 W	6.0 A, 960 W	32.2	R
244507	3/4	230	4“ 2-wire	6.8 A, 940 W	8.0 A, 1310 W	40.7	N
244508	1	230	4“ 2-wire	8.2 A, 1210 W	10.4 A, 1600 W	48.7	N
244309	1.5	230	4“ 2-wire	10.6 A, 1710 W	13.1 A, 2280 W	66.2	M
224301	2	230	4“ 3-wire	2025 W	2555 W	53.1	G

Franklin’s own note sits directly beneath that table and travels with every figure taken from it: “Performance is typical, not guaranteed, at specified voltages and capacitor values. Performance at voltage ratings not shown is similar, except amps vary inversely with voltage.”

The immediate lesson for anyone who has been budgeting backup power around a well pump is that the running figure is smaller than the folklore. A 1 HP submersible motor at 1210 watts full load is in the same neighbourhood as a hair dryer. That is why the wattage question, on its own, has always made a well pump look reachable.

The 115 volt exception, and why it has the highest starting current

Do not carry away from that table the belief that well pumps are 230 volt devices. Franklin publishes a 115 volt option, and it is worth reading closely.

Model prefix 244504 is a 1/2 HP 115 volt 2-wire motor. Its running watts are identical to its 230 volt sibling at 670 W full load and 960 W maximum, because the power is the same either way. What changes is current, and it changes in both directions at once: full load amps double from 5.0 to 10.0, and locked rotor amps double from 32.2 to 64.4.

Reading down the locked rotor column, that 64.4 is higher than every other motor in Table 29 rated 1 horsepower or less, including the 1 HP 230 volt models at 48.7 and 41.8. It is not the largest figure in the table, which belongs to the motors above 1.5 HP, but it is the largest one attached to the smallest amount of work. That comparison is ours, read off Franklin’s column rather than stated by Franklin.

The reason is arithmetic that Franklin’s own note points at. Halve the voltage and, for the same power, the amps double. So the one residential configuration where the voltage objection below does not apply is also the configuration presenting the largest inrush current per horsepower.

Two tables in one manual that do not agree, and that is not an error

This is why a bare figure for “a 1 HP well pump” is not publishable and why every number above carries a table and a model prefix.

Franklin’s Table 30 covers the same nominal horsepowers for motors with a “G” suffix in the model number, and its figures are different. The 1 HP 230 volt 2-wire model reads 1210 W full load, 1600 W maximum and 48.7 locked rotor amps in Table 29, and 1215 W, 1630 W and 47.6 locked rotor amps in Table 30. The 1/2 HP 230 volt 2-wire model reads 670 W and 32.2 locked rotor amps in Table 29, and 660 W and 28.5 in Table 30. Same manufacturer, same manual, same horsepower, different motors.

Neither table is wrong. They describe different products. What that means for a reader is the same thing our sizing worksheet has always said: the figure that governs your decision is the one printed on your unit, and horsepower alone does not identify a motor.

Locked rotor amps are amps, not watts

One unit caution, because both of the power pages on this site already commit us to it and it is very easy to get wrong in exactly this context.

Locked rotor amps is a current figure. Multiplying it by the supply voltage gives volt-amps, which is apparent power, and apparent power is not watts. At the moment a motor starts, its power factor is low, so the volt-amp figure is meaningfully larger than the real power drawn. Comparing an LRA-derived volt-amp figure against a generator’s or a power station’s watt rating is comparing two different quantities and will produce a wrong answer in either direction depending on which you inflate.

Franklin sidesteps this for you by publishing a separate table meant for exactly this purpose, in the units generators are actually rated in, which is the next section.

What Franklin Says a Generator Has to Be

The AIM Manual has a section headed Engine Driven Generators, and it is the motor manufacturer’s own answer to the question this page exists to settle.

It opens with the constraint that makes pump starting different from everything else on a backup power list: “Generators must be sized to deliver at least 65% of the motor’s rated nameplate voltage during starting to ensure adequate starting torque.” A generator that sags below that during inrush does not merely start the pump slowly. It fails to develop the torque to start it at all.

Then Table 3, headed Generator Minimum Rating, with this note above it: “This table applies to 2–wire, 3-wire, or 3-phase motors. For best starting of 2-wire motors, the minimum generator rating is 50% higher than shown.”

Motor rating (Franklin AIM 2023, Table 3)	Minimum generator rating, KW	Minimum generator rating, kVA
1/3 HP	1.5	1.9
1/2 HP	2	2.5
3/4 HP	3	3.8
1 HP	4	5.0
1.5 HP	5	6.25
2 HP	7.5	9.4

Applying Franklin’s own 50 percent note to the 2-wire residential motors in Table 29 is arithmetic, and the arithmetic is this site’s rather than Franklin’s, so here it is shown rather than asserted: a 1/2 HP 2-wire motor goes from 2 KW to 3 KW, a 3/4 HP from 3 KW to 4.5 KW, a 1 HP from 4 KW to 6 KW, and a 1.5 HP from 5 KW to 7.5 KW. Whether that uplift applies to your own pump depends on whether its motor is 2-wire or 3-wire, which Table 29 distinguishes by model prefix and horsepower alone does not tell you.

Three further instructions from that section, quoted because they are safety statements and because the third of them prevents a specific mistake a reader might otherwise make after reading this page.

On connection: “To prevent accidental electrocution, automatic or manual transfer switches must be used any time a generator is used as standby or back up on power lines. Contact power company for use and approval.”

On sequence: “Always start the generator before turning on the motor and always turn off the motor before shutting down the generator.” Franklin’s reason is that the motor’s thrust bearing may be damaged if the generator coasts down with the motor running or runs out of fuel, which is a real risk in an outage where refuelling happens in the dark.

On soft starters, and this is the one to read carefully if you have been shopping for a device that promises to let a smaller generator start a pump: “Do not used a Reduced Voltage Starter with a minimum sized generator.” The wording “Do not used” is Franklin’s own typographical error, reproduced exactly. Its explanation, outside the quotation marks because the printed sentence carries a line-break hyphenation and a second typo, is that both the starter and a minimum-sized generator reduce output voltage, and combining them risks severe motor damage and failure from low voltage. Franklin adds a separate note that a minimum-rated generator already acts as a soft start to the motor and that no additional voltage reduction is allowed.

So devices that reduce a pump’s inrush do exist and Franklin addresses them. What Franklin does not do is treat one as a substitute for adequate generator capacity. That is the scope of the negative this page is willing to state.

One boundary on the whole section. Franklin’s generator guidance is written about engine driven generators with voltage regulators, and it says so in its own heading. It does not address battery inverters or portable power stations, and we are not going to extend it to them, because a battery inverter’s behaviour under inrush is a different question that this manual does not answer.

Set Against the Backup Power This Site Recommends

Our power station roundup compares five battery stations for pet loads, and our generator versus power station page uses the Honda EU2200i as its quiet-generator benchmark and names a well pump as one of the loads where a fuel generator earns its place rather than a battery. Franklin’s tables confirm the category call, which is an engine driven generator rather than a battery. What they do not confirm is that particular machine, and the reason is what this section is about.

The voltage finding

Every 4-inch single-phase model in Franklin’s Table 29 from 3/4 HP upward is a 230 volt motor. Only the 1/2 HP appears in that table with a 115 volt option.

Against that, here is what the manufacturers of the machines this site recommends publish about their own AC output, each read on their own product page on August 15, 2026.

Product (manufacturer’s own published AC output)	Rated AC output	Surge	Voltage as published
Honda EU2200i inverter generator	1800 W rated (15 A)	2200 W max. (18.3 A)	“120V 2200W max. (18.3A), 1800W rated (15A)”; receptacles listed as 20A 125V Duplex
Jackery Explorer 2000 Plus	3000 W max	6000 W surge peak	“(4x) 120V~ 60Hz, 20A Max (1x) 120V~ 60Hz, 25A Max”; 120/240V described as expandable voltage through parallel connection
EcoFlow RIVER 2 Pro	800 W total	1600 W	“Pure Sine Wave, 800W total (surge 1600W), 120V ~ 50Hz/60Hz”
Anker SOLIX C300	300 W rated	600 W with SurgePad	Wattage published in Anker’s own FAQ; the specification set behind the product page gives AC Output Voltage as 120V

The finding is categorical, and it is about these products rather than about their categories: none of the machines named above outputs 240 volts as a single unit, so none of them can supply a 230 volt motor, at any wattage. Jackery is explicit that 240 volt output on the Explorer 2000 Plus is an expandable capability reached by connecting units in parallel rather than a property of one unit, and Honda’s own model descriptor for the EU2200i is a 2200 watt 120 volt inverter generator.

That is a different kind of “no” from the one people expect. A 230 volt pump is not a big load. It is a load your 120 volt machine cannot reach by getting bigger.

The wattage finding, for the one case where voltage is not the answer

For the 1/2 HP 115 volt motor, Franklin’s 244504, the voltage objection does not apply, so the wattage question becomes the live one. Franklin’s Table 3 puts the minimum generator for a 1/2 HP motor at 2 KW, and its note raises that by 50 percent for 2-wire motors, giving 3 KW for the 2-wire version.

Set against that: the Honda EU2200i publishes 1800 W rated and 2200 W maximum, the Anker SOLIX C300 publishes 300 W rated, and the EcoFlow RIVER 2 Pro publishes 800 W total. All three are below Franklin’s 3 KW figure for the smallest residential motor in its table, and the only one of them that reaches even the 2 KW figure before the 2-wire adjustment is the Honda, and only on its 2200 W maximum rather than its 1800 W rated output. The Jackery Explorer 2000 Plus at 3000 W maximum is the only one that meets the 3 KW number, and meeting a manufacturer’s stated minimum is not the same as a manufacturer saying the combination will work.

We are not going to tell you that any of these will start any pump. Franklin’s generator guidance is written about engine driven generators, its 65 percent voltage requirement is a behaviour under load that no product page above publishes, and the decision belongs to a licensed well or electrical contractor who can see your actual motor, your actual wire run and your actual machine. What the numbers settle is the easier half: for a 230 volt pump the answer is no on voltage grounds, and for the 115 volt exception the answer is a question for a professional rather than a purchase you make from a comparison table.

What this section does not say

It does not say that no power station can run a well pump. It says that the specific products this site recommends, at their published outputs, cannot supply a 230 volt motor. Larger and split-phase equipment exists, including Jackery’s own parallel configuration, and this page has not evaluated it.

It does not say a generator is the wrong tool. Franklin’s own Table 3 is a generator table, and NEHA’s factsheet suggests connecting the pump to one: “Consider connecting your well pump to a generator if you experience frequent power outages.” PrivateWellClass says the same in slightly different words, recommending a backup power source such as a generator or solar panels connected to the pump for rural or outage-prone areas.

And it does not say anything about how a generator should be run around animals. Carbon monoxide, placement distance and the alarm requirements are covered in full on our generators, carbon monoxide and pets page, and nothing on this page changes any of it.

So the Answer Is Water You Stored Before the Outage

The outage documents above converge on the same instruction, arriving from four different directions. Michigan tells you not to draw the system. Ohio tells you the treatment is off. NEHA and PrivateWellClass tell you to use bottled or stored water. Franklin tells you what it would take to run the pump instead, and it is a contractor-scale answer rather than a shopping one.

The three-day figure, at its real scope

NEHA’s Private Water Network factsheet, under the heading Before a Power Outage and the subheading Essential Steps to Take, gives the only stored-water duration we found in any of these documents: “Buy or store safe water, such as bottled water, to use for drinking, cooking, bathing, and flushing toilets for three days.”

Read what that sentence names. Drinking, cooking, bathing, flushing toilets. It does not name an animal, and neither does anything else in that document except an instruction to direct flushing water away from animal habitats. Extending a three-day human figure to your pets is this site’s own extension of guidance written for people, not something NEHA published, and we are flagging it rather than quietly inheriting it.

Michigan’s temporary supply list adds a distinction worth copying: “Keep a supply of bottled water on hand as emergency drinking water. A few 5-gallon pails filled with water and stored in your basement are helpful for toilet flushes and other nonpotable uses during power outages.” Two grades of stored water, with the drinking grade kept separate. On a well that separation is more useful than it looks, because the flushing water is what protects the drinking water from being spent on a toilet.

Sizing is a different page’s job

How much water a specific household of animals actually needs per day is a real question with a contested answer, and it is not this page’s question. Our page on how much water per dog in an emergency works through where the commonly cited per-pound rate comes from, which bodies publish a day count instead of a rate, where the prep sources disagree with each other, and what the container criteria are. Take your number from there, then read this page for the reason the well household cannot improvise its way out of getting that number wrong.

The timing problem a well household has and a municipal one does not

This is a small point with a large consequence, and it is this site’s own reasoning drawn from two facts rather than an instruction any source wrote.

Standard outage advice includes filling containers when the power goes out, and on a municipal system that works, because municipal pressure does not depend on your electricity. On a well it does not work, for the reason Michigan states: opening faucets during the outage is the thing that depressurizes the system and raises the contamination risk. A well household’s window for filling containers closes before the outage starts, not after.

Which means the trigger for topping up stored water is different too. It is not the lights going out. It is the forecast, the storm track, or a utility notification. If you are in an area with scheduled shutoffs or staged grid alerts, our pets and power outages hub has a section on which outage clock you are actually on, and the notification is the lead time you spend on exactly this.

The Pet Loads That Quietly Assume a Tap

Bowls are the obvious one, and they are not the only one. A pet household has several water demands that outage guidance treats as trivial because, on municipal water, they are.

The aquarium. Our aquarium and reptile power outage guide covers what a tank needs during an outage in detail, including the ammonia threshold that triggers a partial water change, the instruction to add warm dechlorinated water when a tank has run cold, and the way salinity climbs in a marine system as water evaporates and salt does not. Every one of those instructions is written around water arriving from somewhere. On a well with no power, it does not.

We are not going to tell you what to substitute. Which water is appropriate for a stocked tank is a chemistry question with a conditioner, a source and a species attached to it, and we did not locate any source addressing what an aquarium keeper on a private well should use when the pump is down. That is a description of the searching we did for this page rather than a claim that no such guidance exists. What we can say is that it is a question worth asking your aquatic specialist before a storm season rather than at hour six, because the tank guidance and the well guidance were written by people who never expected to be read together.

The fountain reservoir. A circulating fountain is a reservoir that empties over days, and it empties faster in warm weather, which is when outages cluster. Our page on what happens to fountains and automatic feeders without power covers the device behaviour. The well-specific part is that the refill you would normally do without thinking is the faucet you have just been told not to open.

Cleanup. An animal that is stressed by an outage may have an accident indoors, and a household without running water has to clean it with something. Michigan’s two-grade storage suggestion is the useful shape here: a drinking-grade supply kept sealed and separate, and a non-potable supply in pails for everything else. On a well, that separation is what stops a night of mopping from eating the water the animals were supposed to drink.

Hand washing. This one is worth naming because it is the step that quietly disappears when the tap does. If the reason your water is in question is that bacteria, viruses and other disease-causing organisms may have entered a depressurized system, then handling bowls, a fountain reservoir or an animal’s waste during that window is handling something that may carry them. None of the well documents we read draws that conclusion, so it is this site’s own reasoning rather than a published instruction: keep a supply of stored water aside for washing hands with soap after handling those things, for as long as the water question is open.

If You Have Already Drawn the Tank

Plenty of people will read this page after the fact, having spent the reserve on a normal evening before anyone thought about it. That is the ordinary case, not a failure.

What the documents say happens next is a testing sequence rather than a treatment. Michigan sets the standard: after a full depressurization, do not consume until testing verifies the water is potable, with two consecutive samples collected 8 hours apart showing no coliform bacteria. NEHA’s version is to have the well water tested at a certified water testing laboratory after a power outage, and to keep drinking, cooking and bathing with bottled or stored water until lab results confirm the water is safe and repairs are complete. Chester County adds the line flush before drinking and notes that disinfection may be advisable.

Three things this page will not do here, and each omission is deliberate.

It will not give you a disinfection procedure. Well disinfection is a contractor-and-health-department procedure with its own concentrations and hold times, and this site is not the place to learn it from. Michigan’s fact sheet names the local health department as the contact for sample bottles, test selection and interpretation, and NEHA names the local health department, a certified laboratory, or a university extension service.

It will not tell you to treat the water at the point of use and give it to an animal. Our page on water purification for pets covers portable treatment gear and is careful to note that the agency disinfection guidance it quotes is written for human drinking water. Nothing found for this page extends any of that to a well that lost pressure.

And it will not tell you what to watch for in an animal that already drank it. That is a veterinary question. What the human-side documents establish is that the hazard class is bacteria, viruses and other disease-causing organisms entering a depressurized system, and that it arrives with no change in taste, odour or clarity. If a pet drank from your system while it was depressurized, or during an outage on a system with a powered treatment stage, call your veterinarian and tell them that specifically, rather than waiting for a symptom to make the call for you. An animal that is vomiting, has diarrhoea, is lethargic, or has stopped drinking is a call to make now, not tomorrow.

When the Power Comes Back

PrivateWellClass publishes the restart sequence, and it opens with a caution about its own scope: the steps may vary by pump type and model, and consulting the manufacturer’s instructions or a professional is the right move if anything is unclear. The steps are to check that power has been fully restored before attempting to start the pump, then to inspect, in its own words: “Take a look at the well pump and the surrounding components to ensure that there are no visible damage or leaks.” The grammar there is PrivateWellClass’s own. Then prime the pump if yours requires priming, following the manufacturer’s instructions for doing so, turn the power back on once the system has been confirmed in good condition, and monitor: “After starting the well pump, closely monitor the system for any unusual noises, leaks, or other issues. If you notice any problems, it is recommended to contact a professional for further assistance.”

NEHA’s recovery section is an inspection list rather than a restart list, and it names the components to look at, including electrical wires and connectors, the well cap, seals and vents, casing and aboveground piping, indoor piping, storage and pressure tanks, and overflow pipes. Its instruction on damaged wiring is unconditional: never touch damaged electrical wires under any circumstances, and contact a qualified electrician.

For the animals specifically, the sequence that follows from the documents is: keep them on stored or bottled water until the testing question is answered, which is NEHA’s own pairing rather than an assumption that you stored any, run the line flush Chester County describes before anyone drinks, and treat the bowls the same way you would treat a drinking glass. There is no separate, laxer standard for a dog’s bowl in anything we read, and inventing one would be exactly the kind of reasoning this site tries not to do.

What We Could Not Reach, and What That Does and Does Not Mean

Several of the documents behind this page fought back, and one of this site’s standing rules is that a failed retrieval gets named rather than converted into a finding. Here is the full status, so you can weigh the figures above knowing how they were obtained.

Honda’s specification page returned an error to our automated request on August 15, 2026. Because the 120 volt figure is this page’s central conclusion, we did not take it from a search summary or from a retailer. It was read directly from Honda’s own model page in a rendered browser session on August 15, 2026, where the AC Output row reads 120V 2200W max. (18.3A), 1800W rated (15A), and the model’s own descriptor is a 2200 watt 120 volt inverter generator.

Amtrol’s specification card took a second route. An automated request did not return it, so the current specification card, form MC7025 (03/24), was obtained from amtrol.com by a different route and then read directly. That matters because our first copy of that form was the 2014 edition on a distributor’s server, and document currency has been a real failure mode on this site. Checking the current edition turned out to be worth doing: the drawdown figures for the models on this page were unchanged, and two other entries were not.

The Ohio Department of Health’s downloadable fact sheet did not resolve for us. The page itself carries the paragraph we quote in its own body text, and that is what we cite. The fuller PDF behind its download button returned an error, so we do not know what else it contains and we are not going to characterise it.

EcoFlow’s standard RIVER 2 Pro product URL returned a not-found page. The specification we quote came from the refurbished listing on EcoFlow’s own US store, which is the URL this site already cites for that model, and which publishes the AC output as 800 W total with a 1600 W surge at 120 V.

Anker’s C300 product page carries its AC output voltage in the page’s own specification data rather than in the copy you can read on screen. The visible page gives the 300 W rated and 600 W SurgePad figures in Anker’s own FAQ block; the specification set behind that page gives AC Output Voltage as 120V. We read both, by two retrieval routes. Its 300 W rated output is far enough below Franklin’s smallest listed generator minimum that the voltage question does not decide anything for it either way.

We did not find a pet-specific well-and-outage document. Every source on this page is a motor manufacturer, a tank manufacturer, a state or county health authority, a CDC-funded professional association, a state water survey, or a land-grant extension service. None of them addresses animals. We searched on August 15, 2026 and did not reach one; a state veterinary association or an extension service may well publish one we did not find. What we have not done is treat that gap as permission to invent the veterinary half.

Your Well-and-Outage Checklist for a Pet Household

• Find out today whether your household is on a well or on municipal water, and write it on the same page as your veterinarian’s number. Half the households that need this page do not know which one they are.
• Read your tank’s model number off its label and look up its drawdown, not its tank volume. On an Amtrol Well-X-Trol the current sheet is form MC7025 (03/24) and the column is headed Drawdown (Gallons).
• Read your pressure switch setting off the gauge near the tank, because the same tank gives roughly 30 percent more usable water at 30/50 than at 50/70.
• Read your pump’s own plate and write down the model number, the horsepower, the volts and the amps. Franklin’s own manual shows two tables in the same document giving different figures for the same horsepower, so horsepower alone does not identify your motor.
• Store the animals’ emergency water before the season, in sealed containers on a rotation schedule, sized from a per-animal daily rate rather than from what is in the tank. On a well you cannot fill containers from the tap once the power is off without doing the thing Michigan EGLE tells you not to do.
• Keep a second, non-drinking water supply separate from the drinking supply. Michigan suggests pails for toilet flushing, and on a well that is what stops the drinking water being spent on a toilet.
• Label the breaker that controls the well pump now, so that turning it off during an outage takes five seconds rather than five minutes with a flashlight.
• During an outage, treat the pressure tank as a clock rather than a supply. Michigan EGLE instructs owners to refrain from opening faucets, taking showers or flushing toilets during extended outages, and states the contamination that follows depressurization can arrive with no change in taste, odour or clarity.
• If the outage has already started and there is no stored water in the house, get water from somewhere the well is not, rather than leaving a bowl empty. Sealed bottled water already indoors, containers someone filled before the outage, and a neighbour, relative or store on municipal water, where the tap does not depend on your electricity, are all off the well. That routing is this site’s own reasoning rather than an instruction any of these documents publishes.
• If your well runs through a powered treatment unit, know it. The Ohio Department of Health states that an unpowered treatment unit will not treat the water, making it unsafe for drinking, which is a hazard independent of pressure.
• If the system lost all pressure, plan on testing rather than on judgement. Michigan’s clearance standard is two consecutive coliform samples collected 8 hours apart, and NEHA says to use a certified laboratory and stay on stored water until results come back.
• Call your veterinarian if an animal drank from a depressurized system or from an untreated powered-treatment system, and say that specifically. Do not wait for a symptom.
• If backup power for the pump is genuinely on your list, take Franklin’s Table 3 figures and your motor’s plate to a licensed well contractor or electrician rather than to a product comparison. The generator and the power stations we checked for this page are published by their own manufacturers as 120 volt units.
• Never connect a generator to your home’s wiring without a transfer switch. Franklin’s manual states that automatic or manual transfer switches must be used any time a generator is used as standby or backup on power lines, and to contact the power company for use and approval.
• Keep your local health department’s number with the veterinarian’s. It is the contact both Michigan EGLE and NEHA name for sample bottles, test selection and result interpretation.

The Version You Do on a Quiet Afternoon

Four things, none of which takes long, and together they turn the whole of the above into a ten-minute event instead of an overnight problem.

Walk to the tank and photograph the label. You want the model number and, if it is on there, the tank volume. Then photograph the pressure gauge so you have the switch setting. Two photographs settle the gallons question permanently.

Walk to the pump’s disconnect or the panel and label the breaker. If your pump is a submersible in a drilled well you will not be reading a plate off the motor itself, but the paperwork from the installation, the pressure switch enclosure or a service sticker on the tank usually carries the model and horsepower. Photograph whatever you find.

Stage the animals’ water where you will actually reach it in the dark, and put it on the same rotation schedule as the rest of your kit. Our kit rotation checklist turns that into dated tasks, and our pet food and water storage page covers the containers.

Then call your local health department once, before you need them, and ask two questions: where sample bottles come from, and what the turnaround is for a coliform test in your county. Michigan’s clearance standard already tells you the sampling takes at least eight hours end to end. The laboratory’s queue is the one number none of the well-and-outage documents on this page publishes as a general figure, and it is the one that decides how long your animals are on stored water. Our pet water after a flooded private well guide carries the one specific account we found anywhere on this site, a well driller’s estimate of about a month during one hurricane’s aftermath in one state, alongside the EPA mobile labs deployed to shorten it. Treat that as a single data point rather than a number to plan against, and ask your own county the same question rather than borrowing someone else’s disaster.

Where to Go Next

This page is the quantified version of one paragraph on our pets and power outages hub, which maps all four household dependencies that genuinely need backup power and covers which outage clock you are on.

If the question is how much water to store. How much water per dog in an emergency works through the competing per-pound rates, the day-count standards, and the container criteria, and it is the page that sizes what this one tells you to stage.

If the question is whether to buy a generator or a battery. Quiet generator versus power station for pets makes that category decision for pet loads, and generator sizing for pet loads is the worksheet for totalling what you actually need to run. If a generator is already part of your setup, read generators, carbon monoxide and pets before you run one near an animal.

If the tap is running but you have been told not to trust it. Pets and a boil water advisory covers the municipal version of this problem, which has different instructions and different published pet guidance.

If the outage is going to be long. Inland hurricane preparedness for pets works through what a multi-week loss of power and water actually demands, using a documented case where clean water took far longer to return than electricity did.

Then do the one thing that pays off most, and it takes two minutes. Go and read the model number off your pressure tank. Everything else on this page is downstream of knowing whether the number you have been imagining is 44 gallons or 11.8.