April 13, 2016
Purpose: There is some disagreement as to whether boiling water can be poured down a residential kitchen sink without damaging the drain pipe. It might be assumed that if the pipe drains quickly, the amount of time necessary to cause damage would be greater than the actual time that the boiling water would be present in any particular section of pipe. Assuming that this theory is correct, there is a rebuttal, that kitchen sinks can become clogged or partially clogged, or that a cumulative effect of dumping boiling water down a drain on a regular basis may (eventually) cause the pipe to fail, or collapse in areas where the pipe is buried. In fact, collapsed pipes are not uncommon in the plumbing industry; however, it is unknown to the author at the time of this writing, whether any published works exist which cite temperature as the cause of the collapsed pipe or whether exceeding the maximum temperature rating (140 °F) for PVC pipe has any significant, real-world consequences. This experiment was designed and conducted to measure the extent and rate of PVC warping when (a drain pipe is) filled with boiling water, and to measure the duration of time necessary for water to cool, to within the acceptable temperature range of PVC pipe.
Materials and Methods: A section of schedule 40 PVC of 1 1/4" nominal dimension was selected, which had been previously used as a kitchen sink drain. More specifically, this pipe was originally part of the plumbing of a sump pump based drain which had been eliminated. The brand name of the pipe is known but has been omitted from this report to avoid any possible disparagements that might be inferred by the reader. The pipe was constructed in a short U shape with 90° elbows on each end; it was selected from used scrap materials for this experiment because the shape could retain water, because it accurately represented residential sink drain material, and because it appeared to be free of any structural defects or deformations. The total length of pipe from the ends of each elbow was 50 inches. A ball valve was also included in the length of pipe centered at exactly 11 1/8" from the end of the elbow associated with the short arm. The long arm of the pipe was 16 3/4 inches tall, measured from the outside bottom of it's respective elbow; the short arm was 7 inches tall, measured from the outside bottom of it's respective elbow. The pipe was weighed and found to be 1558.5 grams. Because the pipe had extra 9 3/4" length in one arm and the other arm had one half of a union fitting this added an amount of weight to the overall pipe, which possibly makes the total measured weight irrelevant for the purpose of accurately calculating heat transfers.
The pipe was suspended at each end by resting the ends on two chairs of equal height, such that the pipe was level. Straps were not used to secure the pipe. The elevation of the pipe was 25" from the floor to the center of the pipe. No external forces were applied; the only forces known to be present were resultant from the weight of the water and pipe, and the strains produced from water at temperatures above, at, and around the maximum rating for PVC (140 °F). The volume of water used was predetermined by using tepid tap water to fill the pipe, and found to be approximatley 1300 ml. The volume inside the pipe was such that the water level was exactly 1" from the top of the short arm of the pipe (or 6 inches high from the outside-bottom of the elbow(s)). It is interesting to note here, that the weight of the water nearly matches the weight of the PVC which contained it (after accounting for the excess lengths of the arms).
A mark was made with indelible marker at the center of the pipe and a camera was used to periodically record and document the amount of sagging that occured over a total period of 30 minutes. A mercury thermometer was inserted into the short arm of the pipe to monitor the change in temperature over time. The experiment was concluded after the measured temperature fell below the maximum rating for the pipe. This was a one-time test, which was not replicated for statistical accuracy. The collected data is reported below.
At 3:36pm, a flask containing 1.4L of boiling tap water was used to transfer approximately 1.3L into the pipe. Boiling water was poured into the longer of the two arms. A thermometer was inserted into the the other, short arm, at the far end of the pipe.
At 0 minutes the mark was 25" above the floor Water Temp = 212 °F; room temperature, and (by default) the temperature of the pipe was 70 °F.
As the liquid was being transferred twisting and warping of the pipe was observed.
At ~1 minute after -0.15625" Temp = 182 °F
At 5 minutes after -0.25" Temp = 176 °F
At 10 minutes after -0.3125" Temp = 166 °F
At 15 minutes after -0.375" Temp = 157 °F
At 18 minutes after -0.40625" Temp = 153 °F
At 20 minutes after -0.375" Temp = 150 °F
At 25 minutes after -0.46875" Temp = 143 °F
At 29 minutes after -0.46875" Temp = 140 °F
At 30 minutes after -0.50" Temp = 138 °F
Results: After 29 minutes the temperature had fallen below 140 °F (the maximum rating for PVC). At 30 minutes the experiment was concluded by emptying the water into another container, in which it was weighed and found to be 1290.1g. Careful measurements were taken to determine that the pipe had twisted approximately 30° clockwise, from end to end (or approximately 7.5° per linear foot). The pipe began twisting and warping as the boiling water was being poured into the pipe. A measurement of the temperature of the water at the far end, at about one minute, shows that the pipe had already absorbed an incredible 30 °F from the (approximately) 1.3L of water. The total sagging was found to be 1/2" inch after 30 minutes.
The greatest amount of deflection was unexpectedly found at approximately 7 inches (toward the center of the pipe) from the center of the ball valve. The maximum deflection was measured to be 7/8 inch (lateral deflection) or a total curvature of about 2.5 inches measured at either end of the pipe. It is also notable that the long arm of the pipe (into which the boiling water was poured, but not where boiling water was present for more than a few seconds, had a deflection of about 3/16 of an inch; the total curvature was 3/4 of an inch as measured at the end of the arm. The depth of the water was measured to be 6 inches from the outside bottom of the elbow(s). With regards to the long arm, the greatest warpage was found above the water line, closer to where the boiling water first entered and made contact with the PVC. The measurements of sagging that were taken periodically, as part of the experiment, were simply vertical measurements of the mark made at the center of the length of pipe. Prior to conducting this experiment, it was expected that the greatest change would be found in the center of the pipe due to sagging; but the unexpected lateral deflection was 75% greater than the vertical sagging; and the actual maximum deflection per linear foot was found at the entrance, where the boiling water was poured into the pipe. A graphical representation of the measured sagging/changes (at the center of the pipe) is provided below.
Conclusion: Obviously the lateral deflection was due to a strain at the joint of the ball valve; the measured values of sagging were likely affected by the twisting and lateral displacement of the pipe. Speculatively, the most probable cause of the lateral deflection was due to a difference in length of the pipe which was concealed by the fitting; in other words, the pipe was probably cut at an angle. It is known that when different materials or different lengths of material have been bonded together, the object will have significant stearic strains when heated, as the two materials will not expand evenly. Consider the following example: length A is 4ft, length B is 4.1ft.; when heated, each material expands 2% in length. So, length A will be 4.080ft and length B will be 4.182. The difference in the (heated) lengths is 0.002ft, which can cause significant "curling" or warping effects.
Further speculations with regard to the cause of observed lateral warping include a difference in temperature absorption at the joint due to an insulating effect, or possibly, latent forces existed from previous usage of the ball valve, which were finally expressed as the pipe became soft enough to allow potential forces to be released (an unwinding or relaxing effect). Speculations like these could be verified or ruled out by futher testing.
Obviously, boiling water can cause deflection in a 1 1/4" (nominal demension) pipe, which was the industry standard for sink drains for many years. It is also fair to assume that the temperature within the pipe is absorbed so rapidly that heating will almost certainly be uneven, resulting in areas that are quickly over heated and more susceptible to failure. Supposing that a pipe was clogged or slowly draining, or perhaps the existence of a cumulative effect of multiple exposures to boiling water, it is reasonable to conclude that pouring boiling water down a drain could cause failure. This would be especially true of pipes that are buried, as pressure from the weight of soil would be present.
In summary, it has been observed here that schedule 40 PVC pipe which has been exposed for less than one minute to temperatures exceeding the maximum temperature rating will deform. This is evidenced by the 3/4 inch warpage found at the area (the long arm) where the boiling water was poured into the pipe; in this area, boiling water only passed through, and did not remain through the duration of the test. The boiling water was only present in the long arm of the pipe for the amount of time necessary to transfer the water, which was approximately 15 to 20 seconds. Also, where pipes are exposed to temperatures above the maximum rating for an extended period of time, they will continue to deform until the temperature dissipates to below the maximum rating. It seems apparent from the graphical illustration above, that the rate or amount of warping nearly parallels the the instantaneous temperature or rate of temperature dissipation.
Discussion: It is important to consider that the amount of water used for this experiment was only about 1.3 liters (0.34 gallons). Often, larger volumes of water are used for cooking, which will necessarily require more time to drain and will likely transfer a proportionally greater amount of heat/energy to a pipe. Also, the length of time necessary for heat to dissipate could be several minutes, or possibly over an hour when larger volumes (like a gallon) of boiling water are poured into a drain, and/or where drain pipes are insulated. The opinion of the Author at this time, is that pouring a whole gallon of boiling water down a kitchen drain would logically have a greater potential for damaging PVC drain pipe than 0.34 gallons which in this experiment, did cause measurable, significant warping, twisting, and sagging. It is also necessary to bear in mind that for proper drainage to occur, drain pipes should have a gentle slope of about 1 inch per 10ft. Since warpage in this pipe was found to be greater than 1/2 inch per foot, it should be obvious that the cumulative effect of warping and sagging is such that boiling water will likely cause improper drainage, which would logically hasten the ultimate failure of PVC drain pipes, because the exposure time in improper/slowly draining pipes will necessarily be greater.
There were some obvious faults with this experiment. Perhaps the most significant difference with respect to a real-world test, is the fact that straps are used to secure drain pipes in residential construction, whereas, no straps were used in this experiment, which allowed the pipe to twist freely. Certainly, proper support would be beneficial for preventing drain failure. Whether the current construction methods, materials, and/or builing codes are sufficient to prevent failure in cases where the the temperature rating for PVC has been exceeded is not known to the Author at this time. Also, because this experiment did not test for a cumulative effect (repeated exposure of boiling water to the same pipe), it was not ascertained whether a cumulative effect actually exists, and more particularly, whether the pipe becomes sensitized or desensitized by repeated exposure. However, strong evidence has been presented here that there is real-world wisdom in avoiding damage potentially caused by overheating a drain pipe.