A Water Leak in the Desert

The homeowners were many weeks away from their La Quinta, California home, escaping the desert heat, when their gardener telephoned them. He said there was water running out from under the front door. They instructed him to turn off the house water and call their security company. Their security company then inspected the outside of the house and also found water coming from the building’s back wall.   They then entered the home and decided the water source was the area around the refrigerator.   After their return, the homeowners used water company records to calculate that the release of water from the inside of the house totaled approximately 8,450 gallons over a period of approximately 5 days.   Thousands of gallons of water plus desert heat triggered the start of an extensive and expensive water damage restoration project.


GEI was assigned to attend a mutual inspection of the extensive flood damage, which was assumed to have originated from the refrigerator and to determine the cause.   Was the cause of the water loss the supply line leading to the refrigerator, the icemaker unit, the water filter, the water lines inside the refrigerator, or the plastic valve that supplied the water to the other components?


Each of the manufacturers of these different components was insured by a different insurance carrier, none of whom wanted to pay for a loss that their insured didn’t cause.  

The main inlet line to the house water supply had been turned off, chained, and locked so no normal water was available in the house.   Regardless of that, water was still slowly trickling from open faucets, showing the shutoff valve to the building did not completely stop the water flow.   The refrigerator was visually inspected.   No cracks or leaks were found on the water filter, and no indication of leakage was found inside the refrigerator, icemaker, or freezer compartments.

The rear bottom panel of the refrigerator was removed and the water valve was inspected.   The black line that ran from the valve up to the icemaker was broken, and there were calcium deposits on the line, which evidenced leakage from this area.   The water valve was removed and water leaked out of the adjacent connectors.

The plastic lines that supplied the icemaker and front-of-the-door water dispenser on the refrigerator have a typical flow rate of approximately 0.5 gallons per minute, as found by measuring the flow rate on similar refrigerators and typical house water pressures.   At the flow rate calculated, it would take approximately 11 days to release 8,000 gallons of water, too slow to account for the massive water bill.

The black plastic line that ran from the refrigerator water valve up to the icemaker was broken, and had a large amount of calcium deposits on the outside of this line, leading to the conclusion that water had been running down the outside of the line for an extended period.   This plastic line was also found to be extremely brittle. During the inspection, a slight bending caused the line to crack and break.   Plastic water lines, in particular the black plastic line that ran from the refrigerator water valve to the icemaker, become brittle primarily due to heat.   The proximity of the plastic line to the hot compressor was a major source of heat, in addition to the desert location of the insured’s residence, which regularly saw ambient heat above 110 degrees F.

A second inspection of the refrigerator was later performed with a full set of tools to individually isolate and test the suspect components.   At this inspection, the refrigerator was plugged into a power source and a water line was connected to the back of the refrigerator, as it would have been on the date of loss.

Once the water was pressurized, a leaking blue plastic valve was evident and documented.   There was a visible split where the valve had failed, and water was observed spraying from it. In addition to visual inspection, a graduated measuring cup was placed below the leak, and four separate trials were performed to document the flow rate from the leak.   The average flow rate leaking from this valve was calculated to be approximately 1.16 gallons per minute.   This flow rate would discharge over 8,000 gallons in a 5 day period.

The leak and the failure that occurred in the plastic valve was due to a manufacturing defect of the component and was not due to high ambient temperatures, or the 10 year age of the valve in question.   The failure was visible as a split in the plastic casing of the valve, which indicated material, manufacture, and/or design problems.


Plastics used in this application undergo cyclical pressure changes, and the vessel must be engineered with adequate wall thickness and proper material selection to prevent failure.   In the actual molding process, most manufacturers will designate the material must be all “virgin” material if failure of the molded component may result in damages or injuries.   In some molding applications, a certain percentage of “regrind” is allowed to be molded.   This “regrind” is a plastic material that has been through the molding process and is reground and added to the “virgin” material to recycle otherwise wasted material.   This practice of using “regrind” reduces the strength of the molded part and is another possible reason this valve failed.


Expert of the Month: Clifford Stover, P.E., M.S.E.

Mr. Stover is a Registered Mechanical Engineer in California. He holds a Bachelor of Science in Mechanical Engineering and a Master of Science in Mechanical Engineering, both from California State Polytechnic University, Pomona, California. He has been a Professor of Mechanical Engineering and Director of the Engineering Project Development Laboratory since 1995 at California State Polytechnic University, Pomona, Ca.   Prof. Stover has 20 years of experience in the design, manufacturing, analysis, testing, and inspection of mechanical systems, has been involved in a wide range of projects including accident reconstruction, failure analysis, patent infringement, and product liability. Prior to his appointment at Cal Poly, Prof. Stover worked as a Design Engineer, Master Machinist, and Auto Mechanic.

An Assortment of Water Leaks

A general contractor built his own home. Rainy weather led to numerous water leaks through the window framing, roofing, and around the porch. He used several subcontractors and was about to sue for faulty workmanship. GEI was called in to inspect the leaking areas and identify and determine the predominant cause and approximate age of the damage seen.


Water leaks were found in the garage roof, the master bedroom balcony, the first-floor level bedroom, the laundry area by the bedroom, the first-floor bathroom, the under stairs closet, and the custom Fleetwood windows.

The garage roof was inspected when the owner rolled back the blue plastic tarpaulin sheeting that was temporarily covering it. The drains from the perimeter of the garage roof showed that the leaf barrier mesh covering had been removed from the drain entry of some of the roof outlets. The drain entry was, in several places, ineffectively sealed with tar around the downspout attachment.


The downspout attached to the drain entry was, in each case, a square section metal downspout connecting to a circular outlet. The upper end of the downspout was poorly manipulated into the discharge from the garage roof perimeter. The garage roof covering did not appear to have any unsealed penetrations by pipes or ducts. However, the inner edge seam by the perimeter gutter appeared not to be sealed around the entire perimeter, thus allowing water penetration into the roof structure.

Poor technical detailing of the seams between the edge of the bituminous felt and the base material of the edge gutter resulted in movement of the felt by expanding and shrinking when subjected to fluctuating temperatures. The resulting flexing of the felt was not matched by the more rigid form of the gutter material, resulting in separation of the felt from the gutter. Following such separations, the tar or other adhesive used was not able to retain its hold on the base material, allowing water to seep between the base material and the underside of the felt when the gutter filled. The edge gutter was very shallow (approximately 1/2 inch to 1 inch), resulting in rapid filling during rainstorms and subsequent movement of water into the gapping between the felt and the base gutter material.  When the drain outlet was clogged to any depth with leaves or other debris, it added to the likelihood that the gutters would remain filled and that there would be continued leakage after a rainstorm. The garage roof covered the bedroom, bathroom, laundry room, and garage.

The master bedroom balcony surface was ceramic tileing. Several of the grouted joints had visible cracking. The outer edge of the balcony had an attached steel railing. The drip edge of the bedroom balcony was coated with mortar and the mortar was cracked. The underside of the balcony had a longitudinal crack with visible water drips when it rained.


The grout placed between the tiles on the balcony had no sealer on the upper surface. This allowed natural cracking through expansion/contraction cycles resulting from heat/cold cycles to produce a multitude of channels for the later rainfall to leak through. Further, the slope of the balcony was inadequate. Water collected and soaked into the grout. Ceramic tiles are impervious to water but the repeated cycles of soaking the cracked grout resulted in water seeping through the grout to the ceiling below.

Another source of leakage was the stucco around the windows. Stucco applied to walls is designed to protect the underlying tarpaper and timber framing from water. When stucco cracks because of earth movement, expansion/contraction, and other related movement of the building envelope, the path for moisture entry is then designed to be resisted by the underlying layer of tarpaper. Tarpaper, when applied around a window, particularly a window that has a sloping head, can be a difficult material to use to ensure complete coverage for a watertight seal. There was cracking to the exterior stucco because of earth movement and flexure of the building envelope. The subsequent exposure to rainfall resulted in moisture seeping through the difficult area of waterproofing to the interior plaster finishes under the stairway.

The drain outlet for the roof above the first floor bathroom was embedded in white sealant mastic. This had cracked and failed to ensure a watertight seal against the downspout. This mastic failure had led to water collecting in the low spot of the drain outlet and rising up to flood into the defective seal between the roof finish and the drain outlet and then to the wall and roof. The construction of the roof allowed water collecting under this drain to move toward the window header and to infiltrate the bathroom window.

The windows in the front entry porch had neoprene gaskets to seal the window glazing into the frame. The gaskets were, in several places, disturbed in their placement, allowing ingress  of water.


The underlying cause of the gasket failure was failure of the seal to be completely applied during the manufacturing process. When the glazing was inserted into the frame, during final assembly, some parts of the gasket were not sufficiently lubricated at the glazing edges, resulting in failure of the gasket to “lip” successfully over the glazing’s edge. Therefore, when a tight fit was not made, water from rainfall ran over the glazing to meet with a lower edge gasket improperly fitted. The water then flowed through the frame construction to the interior of the window unit.

Based on the foregoing, the approximate age of the damages noted was at least six months prior to the date of our inspection. With our report in hand, the homeowner was now well prepared for his lawsuits.