Utility Pumps

Shock or Tingle Near a Utility Pump: What’s Actually Wrong

Diagram of a basement utility pump pit showing one voltmeter lead touching the wet pump housing or water surface and the other lead touching a verified grounding point, with the test run twice, once with the main breaker on and once with it pulled, and the two outcomes labeled: voltage drops toward zero with the main off meaning the fault is downstream inside the house, versus voltage persists with the main off meaning the fault is upstream on the utility side
On this page
  1. If the Basement Is Actually Flooded, Stop Here First
  2. A Felt Shock Doesn't Mean Your GFCI Is Broken
  3. The One Test That Tells You Whose Fault This Is
  4. When the Reading Climbs With Other Loads, It's a Loose Connection at the Panel
  5. When the Reading Doesn't Change, Check What Bonds the Pipe
  6. When It's Actually the Pump
  7. This Isn't a Panel You Open Yourself
  8. The Pump Was Never the Prime Suspect
  9. FAQ

Last Updated on October 11, 2026 by umar bwn

Two very different emergencies get called the same thing here. A fully flooded basement means assume every surface is live and cut power from outside it before touching anything. A utility pump running normally with a tingle felt in the water around it is a different, more common problem — and it’s rarely the pump itself. Most of the time it’s a wiring fault elsewhere in the house, or upstream of it entirely, and one voltmeter test tells you which.

This article is about the second case: the pump is running, it’s moving water, and you just felt something you shouldn’t have while your hand was in it.

If the Basement Is Actually Flooded, Stop Here First

If the water has risen past a few inches — covering outlets, appliance bases, or more of the floor than just the pump’s immediate footprint — none of the diagnostics below apply yet. Treat every surface as energized. Don’t wade in, don’t reach for the panel through standing water, and don’t unplug anything by hand. Get people and pets out, cut power from a dry location if you safely can (an exterior disconnect or a dry stairway to the panel), and if you can’t reach power safely, call the utility to pull the meter before anyone goes back in. That’s a different article’s job to walk through in full; this one picks up once the water is back to a manageable level and the pump itself is running.

A Felt Shock Doesn’t Mean Your GFCI Is Broken

If you’ve already checked and the GFCI protecting that outlet hasn’t tripped, it’s tempting to conclude the shock wasn’t real, or that the GFCI has failed. Neither is a safe assumption, and the reason comes straight from the standard the device is built to.

A Class A GFCI — what’s been required in US dwellings for this kind of location since the 2008 NEC, and codified generally at NEC 210.8(A)(5) for unfinished basement receptacles — is built around two numbers in UL 943. It must trip at 6 mA or above. It must not trip below 4 mA, a floor UL built in, as IAEI Magazine explains quoting the standard’s own rationale, “to prevent unwanted tripping due to the natural leakage in appliances, tools and other connected loads, and in the wiring itself.”

Separately, the current a person can actually feel starts far lower than that. The generally cited human-perception threshold for AC current is around 1 mA — a faint tingle, well below where a correctly functioning GFCI is required to do anything at all. So a real, perceptible shock can exist entirely inside the range a working Class A GFCI is designed to ignore. The device isn’t malfunctioning. It’s doing exactly what UL 943 tells it to do, and that range doesn’t include “zero.” Treat the sensation as real information regardless of what the GFCI did or didn’t do.

This is also a different complaint from a GFCI that trips over and over. If yours is cutting power repeatedly rather than quietly not tripping at all, that’s a separate, already-diagnosed fault — see utility pump tripping the breaker for that investigation instead of this one.

The One Test That Tells You Whose Fault This Is

Before touching the pump, the panel, or anything else, one measurement separates three very different problems: something in the house wiring, something in the plumbing’s grounding path, or something upstream of the house entirely. You need a basic AC voltmeter (a clamp meter or multimeter set to AC volts works) — not a non-contact pen-style tester, which is built to sense field strength near a live conductor, not to measure a small voltage sitting on a wet surface, and gives an unreliable reading either way for this specific job.

Diagram of a basement utility pump pit showing one voltmeter lead touching the wet pump housing or water surface and the other lead touching a verified grounding point, with the test run twice — once with the main breaker on, once with the main breaker pulled — and the two possible outcomes labeled: voltage drops toward zero with the main off (fault is downstream, inside the house) versus voltage persists with the main off (fault is upstream, on the utility side)

One-minute test: with everything running normally, touch one meter lead to the wet pump housing or the water surface itself, and the other to a verified ground — a cold-water pipe you’ve confirmed is actually bonded, or a ground rod clamp, not just any metal you assume is grounded. Note the reading. Then go to the main panel and pull the main breaker (not just the pump’s own circuit breaker) so the whole house is dead, and read again.

  • If the voltage drops to near zero with the main pulled, the fault is somewhere inside the house’s own wiring or bonding, downstream of the main. Keep reading below.
  • If the voltage persists with the main breaker off, the fault is upstream of your house entirely — on the utility’s side of the meter. That’s not an electrician’s job to start with. Call the utility first.

When the Reading Climbs With Other Loads, It’s a Loose Connection at the Panel

If the voltage is present with the main on and disappears with it pulled, the next question is whether it changes when unrelated large loads elsewhere in the house switch on or off — a microwave, a dryer, an AC compressor, anything on a different circuit entirely. If it does, that’s a strong, specific tell for a loose or partially open neutral connection at the main panel or the service entrance, not anything wrong with the pump or its own circuit.

This isn’t theoretical. On the Mike Holt electricians’ forum, a technician diagnosed a tingling outdoor faucet this way: a baseline reading of about 1.5 V between the faucet and moist soil climbed to 10 V when a kitchen microwave was switched on, and to 18 V when a bedroom microwave on a different circuit was added. The voltage was tracking unrelated household loads that had nothing to do with the faucet, which is exactly the signature of a loose connection upstream sharing current across paths it shouldn’t. The fault traced back to loose neutral and ground connections at the panel.

One-minute test: with your meter still connected as above and the main breaker back on, have someone cycle a large appliance on an entirely different circuit — a microwave, an electric dryer, a window AC unit — while you watch the reading. A reading that rises and falls with that unrelated load is a panel-level neutral problem, not a pump problem, no matter how close the pump is to where you feel it.

When the Reading Doesn’t Change, Check What Bonds the Pipe

If the voltage is present with the main on, gone with the main off, but doesn’t move when you cycle other loads, look instead at whether the house’s metal water piping is still bonded the way code requires. NEC 250.104(A) requires a metal water piping system to be bonded to the service equipment enclosure, the grounded conductor, or the grounding electrode system, sized per NEC Table 250.66. That bond is what keeps the pipe at the same electrical potential as everything else grounded in the house. Break it, and the pipe — and anything wet that’s in contact with it, including a pump pit — can sit at a different potential than the ground you’re standing on.

This bond gets broken more often by plumbing work than by anything electrical. A PEX repipe replaces metal pipe (which was carrying the bond) with plastic (which can’t), and unless someone adds a jumper around the new section, continuity is gone. A new water meter installed with a dielectric or all-brass fitting can do the same thing at the meter instead of inside the wall. One documented case traced current felt in sump water to exactly this: a broken bond after a PEX repipe, resolved by installing a new jumper (reported at #6 AWG for a 100 A service, which lines up with NEC Table 250.66’s sizing for that service size) to restore continuity between the metal piping and the grounding electrode system.

One-minute test: with the meter still set to continuity or low-ohms, touch one lead to a metal pipe near the pump pit and the other to a point you know is solidly grounded, such as the panel’s own grounding bar with the panel open and power off, or a visibly intact jumper elsewhere on the piping. A clean, near-zero reading means the bond is intact and this isn’t your cause. An open circuit or a high reading means the bonding path is broken somewhere between those two points — worth mentioning by name when you call an electrician, since it points them straight at the fix instead of a general troubleshoot.

When It’s Actually the Pump

This is the least common of the four, but it’s the one place where the pump itself, not the house, is genuinely at fault — and the one place a replacement is the right answer. Submersible pump manuals across brands (Myers, Sta-Rite, AMT, and others carry some version of the same line) warn against handling the pump or its cord with wet hands or while standing on a wet surface, precisely because a pump with a compromised cord jacket or a cracked motor housing can put stray current into the water it’s sitting in.

One-minute test: unplug the pump, dry it off, and look it over. Run your fingers along the full length of the cord feeling for nicks, flat spots, or soft spongy sections in the jacket — not just at the plug end. Check the housing itself for hairline cracks, especially around the cord’s strain-relief fitting, where flexing concentrates stress over years of use. Any of those found, the pump’s insulation can no longer be trusted, and no amount of rewiring elsewhere in the house fixes it. At that point the best utility pumps for a flooded basement is where to look for a replacement, rather than troubleshooting a sound unit further.

Where the pump sits in the pit matters here too. A pump set directly on bare, silty ground or resting against a cracked foundation wall wears its housing and cord faster than one kept clear of debris and sediment — where to place a utility pump in standing water covers the setup that extends a pump’s working life instead of grinding away at it.

This Isn’t a Panel You Open Yourself

Everything above the last section is safe to do with a meter: touching leads to a wet surface and a verified ground, cycling an appliance, checking continuity on exposed pipe. Opening the panel to inspect or re-torque a neutral connection, running a new bonding jumper, or doing anything inside the meter base is not. A loose neutral at the panel is itself a shock and fire hazard to whoever opens that panel next, and restoring a bonding jumper correctly sized to NEC Table 250.66 is exactly the kind of fix a local inspector will want signed off, not guessed at. Bring your voltmeter readings — both the main-on and main-off numbers, and whether it changed with other loads — to the electrician you call. That alone tells them more than “I felt a tingle” does, and usually saves a service visit.

The Pump Was Never the Prime Suspect

A utility pump sitting in a wet pit is simply where a hand most often completes a circuit that started somewhere else — at a loose panel connection, a pipe that lost its bond during a plumbing job, or a utility conductor outside the house entirely. The pump gets blamed because it’s the thing in the water, not because it’s usually the thing at fault. Running the voltmeter test above before replacing anything is the difference between fixing the actual problem once and buying a new pump that sits in the exact same compromised water next season. For how the pump fits into the rest of the system around it, the utility and transfer pumps guide is the wider reference this article sits under.

FAQ

Is it normal to feel any shock or tingle from water near a utility pump?

No. Every source on this — manufacturer manuals, electricians, utility safety material — treats any felt shock from plumbing or pump water as a real fault needing a fix, not background noise to tolerate. Run the voltmeter test above rather than waiting to see if it happens again.

Can a GFCI fail without tripping, even though I felt a shock?

Not necessarily failing — a correctly working Class A GFCI is built to not trip below 4 mA by UL 943’s own design, while a person can feel current at roughly a quarter of that. A real shock and a GFCI doing exactly what it’s supposed to do aren’t contradictory.

Will a non-contact voltage tester find this?

Not reliably. Pen-style non-contact testers sense the electric field around a live conductor and are meant for checking whether a wire is live, not for measuring a small voltage present on a wet surface. Use an actual AC voltmeter with two leads for this test.

Does a PEX repipe really cause this months or years later?

Yes, if whoever did the work didn’t add a bonding jumper around the plastic section. Metal pipe used to carry the house’s water-piping bond; PEX can’t. The gap can sit unnoticed until something wet — a pump pit included — ends up at a different potential than the ground around it.

More on Utility Pumps