why is water flowing backwards through a 006-B4?

Tomato

This system has worked well for several years.

The pump is installed to reheat water in a buffer tank. There is a check valve between the DCW and the pump. For years it has worked correctly but for the last month DCW has been going past the check valve and pump into the DHW piping after the pump turns on.

The pump and valves are 3/4 copper. The remaining pipe is 1/2 copper. (I realize that the check valve should be 1/2.)

The system was not installed with a way to test the check valve, which is my suspected failure point.

I realize that the pump is not positive displacement, and with enough pressure water will flow the wrong way. I wish I could find an inexpensive, suitable (quiet), positive displacement pump so the check valve would be unnecessary.

Larry Weingarten

Hello, What kind of check valve do you have? Is it a swing or a spring check? Also, how hard is the water?

Yours, Larry

EdTheHeaterMan

I'm not sure how the Taco wet rotor pump works... I'm not a motor expert... Is it possible that the failed check is allowing the reverse flow and spinning the impeller in the wrong direction? ...then when the pump motor is powered it continues to operate in the reverse direction? I don't think the pump will cause a lot of flow operating in reverse, but it certainly won't pump anything in the correct direction!

EdTheHeaterMan

If it worked for a significant time, then just change out (or clean) the check valve and see if that corrects the issue.

Jamie Hall

The Taco is, unless I'm mightily surprised, a centrifugal pump. When the pump is unpowered water can flow backwards through it quite happily. Well, the water is happy. The pump seals may not be... If powered, however, it will try to pump in the correct direction, regardless of which way it is spinning. If it is spinning in the correct direction, all is well. In the reverse direction, it is wildly inefficient. If the pressure in the outlet line from other source is greater than the inlet plus the head added by the pump, water will flow backwards through the pump.

I'm assuming that your drawing is complete: there are no other connections within the loop of the boiler, buffer, and pump than what are shown.

Now. That being said, in your diagram the pump is shown pumping from the hot water line to the cold water line. You state that the flow, however, is in the opposing direction, from cold to hot. This can happen -- in fact will only happen -- if the pressure in the cold water line is greater than the pressure in the hot water line plus whatever the shutoff head of the pump is (and for the little Taco, it isn't that much).

Interesting. I would imagine, though you don't mention it, that this only happens when there is a draw on the hot water line. The flow resistance which is creating the pressure difference causing the reverse flow has to be in the boiler -- not much else. When there is no draw on the hot water, there is correspondingly no flow in the cold supply, and hence the only flow in the system is from the pump -- which will be in the correct direction. As soon as you draw hot water, there is increased flow through the boiler and hence increase pressure drop -- and your poor little pump can't overcome that.

So.

The root cause of your problem is the pressure drop in the boiler -- which probably means that the boiler needs to be descaled or at the very least thoroughly cleaned on the domestic water side. It may be -- probably is -- that the installer anticipated this and installed the check valve to prevent wrong way flow, accepting that the pump will then be operating in a shut off condition (not the best idea for long life, but...). And the proximate cause of your problem is a check valve which isn't working.

Therefore. Fix the check valve and clean and descale the boiler.

Positive displacement, or close to it, do exist, and even for reasonable prices. The problem with them in an application such as this is that if they encounter a pressure difference greater than that for which they are designed, they fail almost instantly in a variety of interesting, but usually permanent, ways.

Tomato

Thanks for the comments.

Larry, It is a spring valve, and the water hardness is "185 PPM (mg/L) or 11 gpg" according to public records. The original check valve(s) were piston but the steel pins holding the pistons in place corroded away; further, the valves were not rated for potable water. However, they were silent, unlike the next one which was a swing. When that one drove everyone nuts with its noise I installed the spring-check. The spring-check is a little quieter.

Ed, Interesting thoughts. I will ask Taco sometime. Perhaps the pdf in the next link will answer the question.

Jamie, Yes, it is a centrifugal pump {pdf linked}; therefore, water can flow past the impellers. There is a little valve on the outlet of the boiler to control flow that I forgot to show. However, I think it will make matters worse if closed. OTOH, it may help preventing cold water "slugs". (Hum, I never thought of that before...)

Regarding the reverse flow: In the text on the bottom of the drawing it mentions that the backwards flow only happens when DHW is being used, so you are correct. Sorry the text is not visible. I don't know how to fix that since it looks OK in MS Edge.

Is there an easy way to determine if the DHW coil in the boiler is scaling? According to Inspectapedia , "If hot water pressure is strong when the water is first turned on (sic) but flow quickly falls off to a weak hot water stream, it's likely that the hot water piping (or tankless coil) are (sic) clogged."

Would moving the temperature sensor to the top of the buffer tank reduce the likelihood of the circulator coming on during a shower?

Additional thoughts: I'm going to install some ball valves and unions around the pump circuit when I replace/clean the check valve (I wish someone would sell a potable water cylinder check valve or a Tesla check valve.)

Today I shut the gate valve and the water temperature stayed tolerable during a shower. That makes me more sure that the reverse flow was the problem with temperatures dropping.

Another question comes to mind... is there a fitting like this? Not a scoop, they work in both directions. This is made to suck liquids and gasses into the flow. I have built a few for vacuum testing and clearing condensate runs. Mine are a little simpler and can be made in a few minutes. I accidently imploded a dome with mine once and regular vacuums did not even seal the bottom gasket. I wish I could replace the tee at the DCW inlet to the boiler where the circulator discharges with something like this.

hot_rod

Caleffi NA 51 is a serviceable spring check, NSF approved for potable water, 1/2- 2” with a selection of connection types

Really no place for a spring check in Hydronics or pumped loops, as you have found out.

Treebeard

Tomato said:

There is a little valve on the outlet of the boiler to control flow that I forgot to show. However, I think it will make matters worse if closed. OTOH, it may help preventing cold water "slugs". (Hum, I never thought of that before...)alt="" />

That is only going to help if it can be opened to allow more flow. The only thing that will help with the "slugs" is matching the shower flow to the boiler's maximum capacity so that the boiler does not shut off while the shower is running. If there are slugs now it is because the shower is not using enough hot water to keep the boiler running.

Treebeard

Jamie Hall said:

Positive displacement, or close to it, do exist, and even for reasonable prices. The problem with them in an application such as this is that if they encounter a pressure difference greater than that for which they are designed, they fail almost instantly in a variety of interesting, but usually permanent, ways.

Seems like a pressure relief valve and bypass pipe could be installed so that there would always be flow if the normal flow was blocked. I think that pressure washers have bypass circuits.

Jamie Hall

Treebeard said:

Jamie Hall said:

Positive displacement, or close to it, do exist, and even for reasonable prices. The problem with them in an application such as this is that if they encounter a pressure difference greater than that for which they are designed, they fail almost instantly in a variety of interesting, but usually permanent, ways.

Seems like a pressure relief valve and bypass pipe could be installed so that there would always be flow if the normal flow was blocked. I think that pressure washers have bypass circuits.

Well yes, of course -- but one might be inclined to wonder why one would go to the expense and complexity of a positive displacement pump when a centrifugal (one moving part) and a check valve (one moving part) will do very nicely, thank you.

Tomato

hot_rod said:

Caleffi NA 51 is a serviceable spring check, NSF approved for potable water, 1/2- 2” with a selection of connection types

Really no place for a spring check in Hydronics or pumped loops, as you have found out.

Are you allowed to recommend a source for the spring check?

hot_rod

Whatever check you decide, as long as it is a hydronic type check valve.
Swing checks are not very good in hydronics, as they need a pressure differential to shut and seal. When a circ shuts off the pressure on the inlet and outlet are the same, there is no force to close the swing. And most are not 100% bubble tight seal.