The Basement That Flooded Only During Long Rain

Why this matters

"It only floods when it rains" is not a symptom. "It floods when it rains for a long time and not when it rains hard" is a diagnosis waiting to be read, because those two conditions point at different mechanisms, and the customer is holding half the evidence before anyone opens a cleanout, the half nobody thinks to ask for, which is the storms that did not flood them. The distinction is intensity against duration, and it is available before anyone opens a cleanout.

Get it backwards and you sell the wrong protection. A sump handles water coming up through the floor and does nothing about sewage coming back up the drain. A backwater valve does the reverse. Both are reasonable products and each is useless against the other's problem.

Stop and contain before anything else

Water on a basement floor with anything energized in it is the hazard that ends this call badly, and it comes first.

• Nobody walks into standing water in a basement while circuits down there are live. If the service panel is upstairs, dry, and reachable without stepping in the water, open the branch breakers serving the basement from there. If the panel is in the basement or the water has reached it, that is a utility or licensed-electrician disconnect from outside, not a wade across the room.
• Before you work on any de-energized circuit or equipment down there yourself, de-energize, lock and tag it and prove it dead with a tester you have checked on a known live source before and after, which is the electrical branch of the isolation rules at 29 CFR 1910.333(b)(2) rather than the mechanical stored-energy standard, and the live-dead-live proving sequence itself is NFPA 70E-2021, 120.5, in the edition your employer's electrical safety program has adopted.
• Treat the water as sewage until you have looked at it. Wastewater contact means splash-rated goggles, gloves and boots, no eating or drinking on site, and washing before anything else, and it changes the cleanup from a wet-vac job to a contaminated-materials job.
• Do not open the floor drain or a cleanout while the level is still rising. A cleanout opened during a surcharge is a hole at the bottom of a column of water, and the water arriving through it comes from the main.

Containment while the event is running is limited and worth stating honestly: get the power off, keep people out, and let it recede. Diagnosis happens after.

The discriminator: intensity or duration

Every wet-basement mechanism is either a rate problem or a storage problem, and rain has both a rate and a total.

• Rate mechanisms overwhelm a path that is sized for a flow. A downspout discharging next to the wall, a blocked area drain, a window well with no drain, an undersized gutter, a yard that grades toward the house. These fail during the hardest fifteen minutes of a storm and stop as soon as the rain eases. Short violent storms cause them; long gentle ones do not.
• Storage mechanisms fail when something fills up. The soil around the foundation saturates and the groundwater table rises, or the area's sanitary system takes on so much infiltration and inflow through defective pipe and improper connections that the main surcharges. These do not care how hard it rained in any one hour; they care how much total water has gone into the ground and how long it has been going in. Long gentle storms cause them; short violent ones often do not.

That is the fork. Ask the customer which storms flooded them and which did not, and you have a shot at the mechanism before you own a shovel.

What each mechanism predicts

The row that does the most work is where the water appears first, and the customer knows it. Water coming up out of a floor drain is not groundwater finding a way in; it is the drainage system running backward.

The record: what the customer knows and what the gauge knows

You need two lists side by side: the dates the basement flooded, and the rainfall for those dates and for the storms that did not flood it. The negative cases carry as much information as the positive ones, and most customers have never been asked for them.

Two conditions on the rainfall figures before they enter any comparison. First, intensity has to be computed over the storm's actual duration, not over the calendar day, and gauges report daily totals, so a storm that ran overnight arrives as two numbers that have to be added back together. Second, the nearest reporting station is not the site, so the depths carry an unknown bias rather than a random spread; use them to rank storms against each other, which the bias does not disturb, rather than as absolute depths at the house.

Worked example: four storms, two floods

The customer's own list, four storms in one season:

Storm C needs the correction stated above applied before it can be used. The station reported it as two daily totals, 1.4 inches and 1.2 inches, because it ran across midnight. Adding the split total back gives 1.4 + 1.2 = 2.6 inches over 30 hours, and without that step storm C enters the table as a 1.4 inch event and looks like the smallest storm of the four instead of the largest.

Now intensity, each over its own duration:

• A: 1.9 / 1.5 = 1.27 inches per hour
• B: 1.1 / 18 = 0.06 inches per hour
• C: 2.6 / 30 = 0.09 inches per hour
• D: 2.2 / 3 = 0.73 inches per hour

Read the three candidate predictors against the two floods:

• Total depth does not predict. The floods were the 1.1 inch storm and the 2.6 inch storm, which are the smallest and the largest of the four.
• Intensity predicts backwards. The two that flooded had the two lowest intensities, 0.06 and 0.09 inches per hour, and the two highest-intensity storms, at 1.27 and 0.73, both left the basement dry.
• Duration predicts cleanly. The two that flooded ran 18 and 30 hours; the two that did not ran 1.5 and 3 hours.

That kills the rate mechanisms outright. A downspout at the wall, a blocked area drain, a window well: all of them fail hardest at 1.27 inches per hour, and storm A did nothing.

Which storage mechanism. The customer reports water first appearing at the basement floor drain and the laundry standpipe, grey and foul, with nothing at the wall and floor joint and a dry sump pit through both events. Against the table above, that is the sanitary surcharge column on every row, not the groundwater column, and the sump pit staying dry through a 30 hour storm is the observation that rules groundwater out rather than merely making it less likely.

Confirming it from outside the house. Two neighbours on the same block reported the same thing on the same two nights. An area-wide response on the same nights is the signature of infiltration and inflow into the public system rather than a defect on one property, and it moves part of the answer to the sewer authority.

What this changes about the recommendation. A sump pump would have run for thirty hours in a dry pit and prevented nothing. The protection that matches the mechanism is a backwater valve on the building sewer, sized and located to the plumbing code as the jurisdiction adopted and amended it, with the understanding stated in writing that it protects against flow returning from the main and does nothing about anything entering from the ground.

And the honest limit: a backwater valve is containment. It closes while the main is surcharged, which means the building cannot discharge either, so every fixture in the house is out of service for the duration of the event. That has to be said out loud at the quote rather than discovered during the next 30 hour storm.

The part of the diagnosis the shop still owns

Area-wide infiltration does not excuse the customer's own lateral. A defective lateral takes on groundwater during exactly the same long storms and contributes to the surcharge it then suffers from, so a wet-weather camera run, done during or immediately after a long rain rather than on a dry Tuesday, is the one inspection that shows infiltration entering. Run it dry and the defects are invisible, which is why most laterals get a clean report.

Look for water entering at joints and at the tap, not for standing water. Water running down the inside of the pipe wall from a joint, with the fixtures upstairs off, is infiltration and it is worth locating with the sonde and writing down.

Where this reasoning breaks

A customer list with only positive cases. If they only remember the floods, duration and intensity cannot be compared and you have no discriminator. Ask for storms that did not flood them, by month if not by date, and accept a rough list.

Two mechanisms at once. A house can have a rate problem at a window well and a surcharge problem at the floor drain, and the four-storm table will look muddled rather than wrong. Split the record by where the water appeared first and run the comparison separately for each location.

A recent change on the property. A new patio, a regraded yard, a downspout rerouted into the ground, or a new area drain tied into the sanitary changes the whole picture and makes the storms before the change non-comparable with the storms after it. Ask what changed and when, and cut the list at that date.

References

• 29 CFR 1910.333(b)(2), electrical safety-related work practices for de-energizing and verifying, general industry; 29 CFR 1926.417 is the construction counterpart for lockout and tagging of circuits
• NFPA 70E-2021, 120.5, in the edition your employer's electrical safety program has adopted, for the live-dead-live proving sequence
• Model plumbing code backwater valve provisions, which bind only as adopted and amended by the local jurisdiction
• Local sewer authority reporting for area-wide wet-weather surcharge events
• See related: What a Cross-Connection Means Inside a Drainage System; Why a Drainage Fault Is Usually a Slope Fault in Disguise