I have a 15-year (and counting) tale of woe concerning a Kohler k304k shower valve. This is a pressure balanced valve body that Kohler has made for many years and may still be producing, consisting of two internal assemblies- the cap, including the control handle, and the pressure balancing unit (PBU). The PBU is the one that looks like a little pair of binoculars. The problem first surfaced several years after the valve was installed.

The crime scene consists of a typical residential installation on city water with hot water supplied by an 80-gallon electric heater. The house is about 40 years old with a typical modern copper water system. The Kohler valve was part of a bath remodeling project and the hot water supply to it includes a gravity operated recirc line back to the heater. This consists of a return line under the supply and a check valve. Being gravity, there is no pump. Hot and cold supplies are ¾" copper. The heater is in the basement and the shower is on the first floor.

This valve randomly goes full cold even with no other demands on the system- faucets, appliances, toilets, etc. This happens sporadically, sometimes not for several years, and sometimes multiple occasions in a week. The PBU is pretty straightforward, with a diaphragm controlling the double piston which regulates the hot and cold sides. When misbehaving, it is apparently acting as if there is a substantial drop in the cold side pressure, shutting down the hot side. As stated, it makes no difference whether any other water outlets in the house are calling for water. And, I can find no other temp fluctuations elsewhere in the house, including another (Delta) shower valve.

Both the cap and PBU have been changed multiple times with scrupulous attention to Kohler's instructions, including the use of the little packet of grease that comes with the replacement parts. A Kohler rep, though very polite, didn't seem to understand anything she couldn't answer with her rehearsed lines. The company is actually sending me two new replacement assemblies, though not requested to, but I'm doubtful that this will be the solution given the history.

I am some years retired from a 40-year career as a high-end building contractor who worked with my people on the job every day. I am sufficiently versed in plumbing to understand how it works, and over the years have done a fair amount of plumbing myself, but I have never come across this before. This issue happens to be in my own house.

The next time I open it up I will flush the supply lines to the valve, although There has never been any grit, debris, etc. in the shower head and no similar problem with any other fixture in the house. Anyone else ever have this issue?

Is the volume of water changing or just the temperature?

Volume does not noticeably change.

If the volume isn't changing, I don't think it's a loss of hot pressure issue, but a loss of hot temperature problem. Can you shut off the return line at the tank? When the shower goes cold how long does it stay cold? Have you checked other faucets for hot water when this happens? If yes, did you keep the shower running? How do you get the hot to resume in the shower?

Steve, thanks for your interest. It has occurred to me as well that this might be a temperature issue. I also have considered that the recirc line might be the culprit, and yes, we did put a ball valve in the line during the original job. But closing it would be inconclusive, as our problem can lie dormant for months or years before acting up multiple times in a day or week. Thus, connecting cause to effect would be problematic.

To answer your question, the recirc line (½" copper) works very well, the flow is continuous, and there is always immediate hot water at any faucet in the bathroom. As you would expect, the line connects to the cold supply into the water heater.

I have just been thinking about what would happen if the check valve (a swing-type check) did in fact get stuck open, and it occurs to me that it would amount to a crossover in which the recirc line would flow backwards, drawing from the feed to the heater and mixing with the hot water flowing in the usual manner where they meet, under the bathroom. And, since the heater end of the recirc line is never cold to the touch, only warm, I could test the check valve by opening any hot faucet in that bathroom and feeling to see if the heater end of the recirc line gets colder. If it does, that's the answer, and the reason previous tests running hot water in the bathroom didn't find it would be that we didn't let it run long enough to let that cold water reach the bathroom (about a 40' run).

If that is the solution, the reverse flow could be happening more often than is apparent, as I tend to take shortish showers and might be done before the temp change arrives at the shower. The supplies to the bathroom, remember, are ¾", carrying more than twice the volume of the ½" return, and water would flow in both lines.

In any case, When I test this theory, I'll post back.

I have a similar setup in my house. I repiped with Uponor Pex all 3/4" with a 1/2" return that goes through a swing check at the bottom of the water heater. When the check was new, no problems. As some crap builds up on it, it doesn't close fully. When I'm in the shower and someone turns on the hot water at the lavy sink, the water goes cooler. If my return line was connected into the cold side of the water heater, then it would definetly be cold. Other than replacing the swing check every once in a while, I have no idea what else could be done for a gravity recuric line.

Problem now solved- Steve (CA) and I reached the same conclusion, that there was an issue with the recirc system. I was able to diagnose cold water backflow in the recirc line when the shower was running by feeling the recirc line temp with my hand on the line.

As I had suspected, the backflow was more than occasional, and the reason it was noticeable only in the shower and not the sinks is that the sinks never ran long enough for the cold backflow to reach the faucets. A faucet would have had to run about 1½ gallons of hot water before that happened.

The check valve is a Y-pattern swing check which has a rubber packing washer in it. The washer was in the process of disintegrating and so it didn't always produce a good seal against backflow. With a new packing, it's all good now. A good lesson to learn, as we have two of these setups in the house, this one and another to the kitchen that was installed several years more recently.

Thanks to Steve and PA_plumber for your responses.



Edited 1 times.

Thanks for posting back, congratulations for finding the problem and glad it's a simple solution.