The scientific method in action determining the cause of a mini-flood

I don’t think many of us would associate the scientific method with a leak in a basement.  In this case, however, a big leak, a mini-flood, in the electrical and modem rooms of a commercial building – not the place you want rogue water.

In chasing down the cause – three possible causes ended up on the table at different times, with the third still there – it struck me I was following the scientific method without thinking.  Nice that it came naturally – it should after a few years investigating failures.

This is how I used the scientific method in this case:

  1. I thought – hypothesized – about the cause of the leak based on the evidence I initially had (see below),
  2. fixed that cause, so I thought,
  3. checked if the rooms were still leaking – a test of a hypothesis in the scientific method -,
  4. saw that they were
  5. so, modified my thoughts on cause – modified the initial hypothesis like in the scientific method,
  6. fixed the modified cause,
  7. checked if …,
  8. etc. etc…

It’s not unlike differential diagnosis in medicine. (Ref. 1) Time consuming and expensive at times but thorough and reliable – the way it must be in engineering investigation and medicine.


I was retained by the owners of a medium sized three story, commercial building to determine the cause of a leak in the electrical and modem rooms.  The rooms served several businesses in the building.  The rooms were separated by a partition and located on the right side of the finished basement next to the concrete walls.

I started my engineering investigation doing standard tasks.

Staff briefed me on the leak problem.  They told me that water appeared on the floor of the rooms during heavy rain and strong winds from the front, right corner of the building, the southeast.  The water flowed across the floor until the storm blew through and the rain stopped.  The source of the water behind the finished walls could not be seen.

The leaking had been going on for several years since building renovations that included electrical services.

I examined the interior layout of the building and also measured the location where the rain water first appeared on the floor.  I referenced the measurements to the rear and right, exterior walls.

I then examined the exterior layout of the building and measured the location of all features on the roof, and also on the right, brick-clad wall where the water was appearing in the two rooms.  The features on the roof and side included roof drains, exhaust pipes from wash rooms, air conditioning units on the roof and an electrical service mast on the right wall.  All places where rain water might get in.

I compared the interior and exterior measurements.  This revealed that the service mast was in the same location on the outside of the building as the mini-flood on the inside.  Good evidence

Electrical service mast construction

The service mast was a 4″ diameter PVC pipe that extended from the roof down to near the bottom of the brick wall and the top of the concrete basement wall.  The service mast carries electric wires/cables from the street to the building.  The mast/pipe was vertical to the bottom of the brick wall then horizontal through a 90 degree elbow into the wall.  The joint between the service mast and the brick wall was caulked.  Fairly standard construction.

The electrical cables from the street entered the service mast at the top through three 1.5″ x 2.5″ holes at the underside of a canopy.  The cables have what is called a “drip” loop so rain water on the cables can drip off before they go into the holes.

I hired a carpenter to take down part of the finished interior wall in the electrical room.  This revealed that the horizontal electrical service mast entered the rooms through a circular hole cut in the top of the concrete basement wall.  The mast then continued along the top of the partition between the electrical and modem rooms.

The hole in the concrete wall appeared to have been drilled several inches too low in 2004.  A plug of concrete was placed between the underside of the mast and the bottom of the misplaced hole to fill the gap.

Rogue water in the wrong place

Water stains on the bottom of the plug of concrete and down the concrete wall indicated the water on the floor was getting into the rooms at the joint/contact between the plug of new concrete and the old basement wall concrete.

Droplets of water and running water were seen at and below the plug of concrete during future rain and wind storms out of the southeast.  Very good evidence.

Based on the evidence I had at this point, January 22, 2020, I thought – my initial Hypothesis #1 – that the leak and the mini-flood was caused by rain water running down the outside of the service mast and into the electrical and modem rooms through inadequate caulking.

I hired the carpenter to apply additional caulking on top of the old and this was done.

Heavy rain and strong winds on February 27 and the floor mini-flooded again.  More evidence – the water is getting in somewhere else; the exterior caulking was not inadequate.

A suspicious gap and “drip” loopless cables

I noticed during the wind and rain storm in February that there was a gap along the steel heading to a window next to the electrical mast and that the steel was rusted.  Could rain soaked up-gusts of wind get into the window gap and mini-flood the floor?  Hypothesis #2.

But before testing Hypothesis #2 by caulking the gap along the window heading I decided to take the back off the elbow in the PVC electrical mast.  This was where the mast changed from a vertical mast to a horizontal mast and on into the electrical room.

I saw that the bottom of the interior of the mast was water stained and covered with dark, damp dust to an estimated 3/4″ up the circular interior of the mast.  The electrical cables were there with caulking everywhere around the cables filling the space between the cables and the inside of the horizontal mast.

Except for an estimated 1/2″ gap or hole at the top of one of the cables between it and the caulking.

The cables did not have a rain water “drip” loop like at the top of the service mast before going into the horizontal section of the mast.

The purpose of the caulking filling the interior of the mast around the cables was to intercept and shed any rain water that might get into the vertical mast and flow down the cables to the caulking.  The water would then flow across the caulking and onto the bottom of the service mast elbow to drain through small holes drilled there for the purpose.

However, this would not happen completely because of the 1/2″ gap at the top of one of the “drip” loopless cables – there’s no caulking at the gap/hole to intercept rain water.

The culprit? Rain soaked up-gusts of wind and “drip” loopless cables?

Some rain water was getting into the holes in the canopy at the top of the service mast.  This was evident by the water stained bottom and damp dust at the bottom of the mast elbow.  I can easily imagine rain soaked up-gusts of wind getting into the canopy holes.

Could enough water get in and some flow down the “drip” loopless cable and through the 1/2″ gap in the caulking at the top of the cable?  Then through tiny holes that might exist in the elbow at the bottom of the mast above the concrete basement wall?  From there into and through the plug of concrete and down the wall to mini-flood the floor below?  I thought so.  Hypothesis #3.

I plan to caulk the 1/2″ gap above the cable when the weather warms up.  The Duct Seal caulking used for this work is firm at room temperatures of 20 to 25 degrees, and difficult to work.  It must be easy to work to ensure adequate caulking of the gap in the tight 4″ diameter space inside the elbow at the bottom of the electrical service mast.

I’ll then wait for the next heavy rain and wind storm – hopefully soon; sorry – to test my third hypothesis.  I’ll let you know the results in an update of this blog.


  1. Blog on Differential diagnosis in medicine and forensic investigation, and soft, initial thoughts on cause.  Posted December 20, 2019



Comments are closed.