need advice for short-cycling water pump

[Chris Pirazzi]

Hi,

The water pump at our place has worked fantastically for more than 3 years with even pressure to all 6 small bungalows regardless of how fully open the tap/shower is.

However, since April, something changed big time and now the pressure seen in every tap/shower pulsates wildly and the pump itself is short-cycling whenever any tap or shower is turned on. The only way to prevent short-cycling and get a steady water stream is to wastefully open 3 taps at once! The pump does not run when no taps are open. The issue is not leaky anything in the bungalows.

When the pump short-cycles it sounds like this, with the motor starting and stopping more than once per second: Click Here To Hear Sound

Depending on how open the tap is, the pump also likes to get in a mode where it cycles 2-3 times rapidly, then runs for about 2 seconds, then repeats, resulting in water that pulses annoyingly for 2 out of every 3 seconds.

For various reasons my landlord has been dragging her feet on this so I am trying to do some DIY diagnosis so that when I finally convince her to do something, it will likely be the right thing and we won't waste time and money.

None of the existing ThaiVisa.com threads on this topic:

• http://www.thaivisa.com/forum/topic/857024-water-pressure-problem/
• http://www.thaivisa.com/forum/topic/400983-question-for-those-with-mitsubishi-ep-series-home-water-pumps/
• http://www.thaivisa.com/forum/topic/886828-water-pump-problems/
• http://www.thaivisa.com/forum/topic/691360-pump-differences/
• http://www.thaivisa.com/forum/topic/183722-problem-with-water-pump-hunting/

are directly relevant to my case because my pump is a Hitachi WM-P 280 GP which has a small accumulator bladder cylinder filled with nitrogen, meaning THERE IS NO SHRAEDER VALVE to re-pump up the lost air pressure, and THERE IS NO LARGE TANK (e.g. Hitachi WT-P series) to drain to re-establish the correct air/water mixture, only the small nitrogen-filled cylinder with a rubber bladder inside and no fill valve.

All the ThaiVisa threads above assume there is either a fill valve, or there is a large tank to drain (e.g. Hitachi WT-P series). We have neither.

Also, unlike many of the ThaiVisa threads, in my case the pump WORKED FINE for many years so this is not a case of having the wrong unit installation, where we needed to install a separate pressure tank from the beginning.

The pump is fed by 12 above-ground concrete-cylinder tanks that are adjacent to the pump (those concrete cylinders are kept full by another pump that is working fine). The water level in the supply tanks is about 2m above the pump.

So the pump is fed by water with a mild positive pressure and THERE IS NO PRIMING ISSUE. I have opened the hopper cap to verify that there are no air bubbles and that water flows freely through the pump when the pump power is off and I open the supply line a bit.

Before and during when I had the hopper cap open, I turned on a faucet in the nearest house and I closed the supply valve to the pump in order to let any small chambers in the pump "drain," just to see if it would make any difference, though I suspect this procedure does not apply to this type of pump but only applies to pumps like the Hitachi WT-P series that have a large tank below the pump. It made no difference anyway.

Please see attached photos for pump configuration.

The closest product for which I could find any marketing materials and manual online is the WM-P250GX2:

• WM-P250GX2 Brochure: http://www.hitachi-th.com/hitachi2014/?act=product_detail&id=11&model_id=9081〈=en
• WM-P250GX2 Manual: http://www.hitachi-th.com/hitachi2014/product/catalog/pdf/WP_WT-P_WM-P_GX2.pdf

Adding to the difficulty, as you can see from the attached pictures below and from the brochure and manual above, the WM-P 280 GP has some kind of special "Life-Extending Pressure Stablilized Unit" gadget right below the pressure switch that is supposed to deliver "constant water pressure," but as a side effect of this they insist in the manual that nobody should ever adjust controls (if any) on the pressure switch. Not sure if this is just job protection nonsense for Hitachi or if there is something legit about this unusual component that means normal solutions of turning the pressure switch adjustments won't work. Both are plausible.

Does anyone here really know what this "Life-Extending Pressure Stablilized Unit" gadget is and does, and what effect it has on diagnosing short-cycling and adjusting the pressure switch?

So, given this data, what are the logical steps for diagnosis?

• When opening the hopper cap I noticed a fair amount of mud, so I am willing to open more chambers to clean out mud but I am not sure which chambers are safe to open without destroying the pump and/or not being able to get the covers on again without special tools. Advice?
• I heard pressure switches often get clogged with grit and stop working. Does anyone know specifically how to access the pressure switch on this unit without damaging the unit (i.e. which screws, etc.; the manual does not have this information)?
  
• I know one possibility is that the pressure bladder cylinder has lost its nitrogen charge and/or the rubber in there has burst. Replacing the whole cylinder costs around 1200B plus a 4 hour trip to the nearest real town, so is there anyway I can be sure if this is the cause? Of course I could also do it Thai Style but I'd rather just throw a bunch of money into the trash.
  
• What else might be causing the short-cycling?

Thanks for all constructive responses.

[Crossy]

It does sound like a flooded pressure tank!

I see your pump has a bladder tank so it's possible the bladder has failed, time for a new tank I'm afraid

Before shelling out for a tank try this:-

Turn off the power and water supply.

Unscrew the tank from its seat.

Shake rattle and roll it until you've got all the water out.

If it has an air valve, pump it up to the recommended pressure (Ah, I see you have none, there may actually be one under that cap, they got the nitrogen in there in the first place).

Re-fit and turn everything on.

[happynthailand]

two thing come to mind:

1- your bladder is shot or the charge on the badder is low

2- do you have a filter between the water tanks and the pump?

if no filter ,remove the pressure switch and check that small pipe is not full of sand or mud

most hardware stores sell bladders(per-charged) and pressure switchs

just make sure the power is off to the pump before you start checking things

had that happen to one of our rent-ell homes,had to replace both bladder and pressure switch, plus install a water filter between the tank and the pump

every thing works fine,have to change the water filter about 3-4 months

[kamalabob2]

The above comments seem sensible and helpful. I might add that it is easy to buy a genuine Hitachi Pressure switch. Yamada would not be my first or second choice of pressure switch for my Hitachi water pump.

[Chris Pirazzi]

Hi,

Thanks for advice so far.

As far as the pressure cylinder, it does not have any valve: instead, the fine folks at Hitachi left behind a brass screw to prevent us from trying to re-fill it ourselves. For pix, see this scary Thai-style DIY post: http://2g.pantip.com/cafe/home/topic/R13107929/R13107929.html

And currently there is no filter between the supply tanks and pump so mud is possibly related (l already saw a fair amount of mud when removing the hopper cap).

The good news is that this morning, the landlord finally broke down and had a plumber come out.

The bad, but not surprising, news is that the plumber just came and did what EVERY Thai plumber does: remove the cover to the pressure switch and screw down the screw to increase pressure on the spring until the pump runs more smoothly, and leave in less than 15 minutes before any of us actually had a chance to test if his genius work was effective on the various taps and showers. In fact his "fix" did not work on my shower (it still pulsed annoyingly) but with a few more turns of the same screw, the shower seems to run smoothly now.

However it seems very unlikely to me that a "loose spring" as he called it is the real problem, especially since the problem began suddenly one day in April, not slowly getting worse and worse. I suspect he has just taken the pump's very limited pressure range caused by some other problem like the cylinder and moved that limited pressure range a bit higher, but there will still be problems.

What do folks think about making that one adjustment? Can this kind of adjustment break the pump in other ways (assuming there is still some other fault)?

I was a bit surprised to see only one screw under the pressure switch cap: I expected to see an adjustment for the pump-on low pressure and pump-off high pressure. Does the screw adjust only one of the pressures? Both pressures at a fixed interval? Does it adjust the ratio of pressures?

Does tightening the screw adjust the pump-on low pressure (or pump-off high pressure) up or down?

Before adjusting the screw at all, the pump only ran continuously when 3 taps were on simultaneously. After tightening a bit, 1 tap could run the pump continuously but the lower-flow shower still pulsed. After tightening the screw further, the low-flow shower could run continuously.

I'm a bit confused here as to what effect tightening the screw had: our pump motor only runs at one speed, so how could adjusting 1 screw make it run continuously into a low-flow shower head that is a tiny fraction of the flow the pump can produce, without breaking the pump? Are we now running the pump in a way that stresses the motor more than designed?

After the plumber's record speed trip and my further turning of the screw, the pulsing so far is gone but it appears that the overall pressure we get from taps is significantly less than what we had before the pump failed in April. Any ideas why?

I am trying to understand the basic principles of operation of the pressure switch so I can make a good guess about whether this adjustment is likely to do more harm than good.

Thanks again.

[rawhod]

Our pump was short cycling on a 10 minute time interval.

I stripped the pump completely and cleaned every thing, refitted it and had the same cycling problem. I have been getting the various bits and pieces together to connect the pressure gauge to one of the outside taps to check the pressure drop during the cycling, thinking that we might have a leak on the system.

Then last week the pump gave up completely.

I got out my multi-meter and checked the power supply, OK. I then checked the power in and out of the pressure switch, power in but not out.

So I dismantled the power switch and discovered that the contacts had been burnt out.

The dreaded black ants had invaded the switch and, I presume, the pressure was leaking out on a short cycle.

So..try changing the pressure switch (750 baht from our local parts shop)

(There are in fact two adjustment screws, one on the very top over the coil spring and one down the side at the end of the flat spring)

They can be seen in one of the photo's here:-

http://2g.pantip.com/cafe/home/topic/R8495623/R8495623.html

Edited May 30, 2016 by rawhod

[Crossy]

See the label on top of the switch?

It's there for a reason

Turning up the pressure, provided the pump still cuts out, is likely stressing your pipework. Oh, and the pump motor.

But since the 'repair' was sanctioned by your landlord, all is good

[rawhod]

The culprit...

[Chris Pirazzi]

Thanks again for your advice; I will certainly check out the pressure switch contacts and look for a name brand part.

Some new data:

With the pressure switch still ridiculously screwed down as earlier today, I turned on just one low-flow shower and verified that the pump (which runs continuously under this setting) becomes too hot to touch after only 5 seconds. I don't remember the pump ever getting this hot, though I'm not 100% sure I felt it when only one shower was running before.

Then I turned on another tap at the same time, and the pump cooled down considerably (even though it is still running continuously).

So this is consistent with the idea that cranking down on that screw is just going to burn out the pump quickly.

I have unscrewed the pressure switch spring screw so that the pump is back to giving us oscillating water pressure at the tap until we can figure this out.

Also, I took off the pressure cylinder from the pump and noted that the cylinder was very full of water (I should have thought to dump all the water into a jug and then compared that volume with the total cylinder volume; oh well) as well as having some coarse dirt (but no more than 5% of the volume of the cylinder). I gently stuck a blunt chopstick into the cylinder and I could not feel anything pushing back on the chopstick, implying that the rubber bladder is popped.

Also, I discovered that there was a 3-4 year old pump of exactly the same model in the shed so I tried its older cylinder on the new pump. I was very optimistic because the old cylinder had a thick, triangle-folded rubber thing just 1cm inside the hole of the cylinder: clearly there was SOME bladderish thing there.

Sadly, when trying the old cylinder on the new pump, the behavior was exactly the same: when running a single shower, different settings of the pressure switch spring would either make the pump oscillate (which is good) but make the water pressure also oscillate at the shower (which is bad), or would make the pump stay on and overheat. And I remember the old pump exhibited the same symptom 3-4 years ago, and the "plumber" at that time also cranked the screw down, and eventually the old pump burned out (hmmmmmmmm).

So, that data suggests that either the old cylinder is bad too (perhaps the rubber bladder is dried solid: when poking it with the chopstick, it won't budge, though I don't want to poke it too hard to puncture it), or perhaps the problem has nothing to do with the pressure cylinder.

However given the nature of the symptom (pump oscillates at low flow (which is good), but pressure also oscillates at the shower (which is bad)), it is hard to imagine any other component that is responsible. If it was the pressure switch that was bad, then either the pump would come on when no taps were open (which it doesn't) or the pump wouldn't switch on and off according to the spring tightness (but it does).

Thoughts?

p.s. if you are wondering why I don't just try replacing some parts anyway, there are three reasons. 1: I want to actually understand what is going on for my own edification and for next time this damn thing breaks, and 2: waste of money, and 3: I have to go on a horrific 4 hour bus to get any quality parts and that is really really really not pleasant so any thinking I can do to avoid 2 trips is worth a million bucks. I live in Pai, Mae Hong Son and although we have a LOT of tourists and ugly pink buildings and hardware stores here, what they stock in the hardware stores is incredibly limited. I will check if they have the cylinder and switch here but I doubt it (they usually only have the parts for the absolute cheapest pumps, and usually crappy copy parts that die in 2 weeks).

[Chris Pirazzi]

FYI for others reading this thread, here is a ThaiVisa thread with some relevant info on pressure switches and one screw vs. 2 screws (though the OP has a different type of pump and a different symptom):

http://www.thaivisa.com/forum/topic/698917-water-pump-pressure-accumulator/page-2

[Chris Pirazzi]

Ordered new pressure tank from CM for 1450B; will update in a few days when it arrives. Probably will replace the switch too since the switch is cheap.

[Chris Pirazzi]

Update.

After much drama involving a friend in Chiang Mai and a minivan courier I got a new pressure bladder cylinder here in Pai.

Good news is that it's the right shape and brand and fitting.

But things in reality can never be clean: the CM shop gave my friend a pressure bladder cylinder whose label shows "charged pressure 0.16 MPa = 1.6 bar == 23PSI" rather than the "charged pressure 0.08 MPa = 0.8 bar == 11.6PSI" of the old cylinder.

First question: should that matter? I'm just trying to work out the theory in my head first. It seems like the upper cut-out pressure of the pressure switch should supersede whatever the max charged pressure of the pressure bladder cylinder is, but maybe I'm not getting the physics fundamentals right here.

I tried the 23 PSI tank anyway with my test case of one low-flow shower tap open. Without adjusting anything else (not even the screw at the top of the pressure switch), the new tank did indeed make the pump cycle much less frequently than yesterday (especially since the old burst cylinder had, after a few days, become completely water-logged and the pump was cycling more than once a second) and made the water at the tap/shower significantly more even.

The pump's duty cycle remained at about 50% on 50% off but ran for maybe 2 seconds per cycle (this seems bad: seems like it should have more % of the cycle off if things are working well with only one low-flow tap on, especially since it is a 275W motor).

However the pump still got too hot to touch when it was running, and the water at the tap/shower is still much less even than before April when all this trouble started. Our gas hot water showers still won't stay lit due to the fluctuating water pressure (they are the on-the-wall instant type and they have many safety cutoffs based on both water and gas pressure, so they can serve as a primitive pressure meter).

Also if I subbed in either of the old pressure bladder cylinders (the water-logged burst cylinder, after emptying out all the water, and the old cylinder from the spare pump where the bladder is visible but seemed possibly hardened up) the behavior of the pump under similar load was about the same as the new cylinder in a 1 minute test period right after screwing the cylinder in.

So even the air-filled burst cylinder behaved the same as a brand new cylinder, at least in the short term before it water-logged again. I guess maybe that suggests that the pump is only using a few PSI of the true capability of the cylinder (presumably the air-filled burst cylinder has a much lower max PSI?). Why would that be?

Tomorrow I will buy a new switch, but here again we have a nasty reality: the only switch available here is not quite the right numbers:

OLD ON: 200 kPa == 2.0 kgf/cm² == 29.0 PSI
OLD OFF: 260 kPa == 2.6 kgf/cm² == 37.7 PSI

NEW ON: SAME
NEW OFF: 240 kPa == 2.4 kgf/cm² == 34 PSI

so the OFF pressure is a bit too low on the new switch. Maybe I can use the screw to make the new switch closer.

I will try to find a pressure gauge tomorrow too so I can tell what is really going on.

I examined the old pressure switch and the contacts are a tiny bit black, but not as much as seen in the picture from rawhod. I did notice that the Philips screw head on the top of the pressure switch is almost completely stripped from Thai geniuses "fixing" it

Any ideas for further debugging?

I wonder if the unit gets hot now because the motor windings have already started to melt from abuse over the past months when the pressure cylinder was burst, and so the thing will always overheat when it runs no matter how many taps are open.

[rawhod]

The upper cut off pressure only sets the static pressure in the system. If you used a pressure switch with a higher upper cut-off pressure, you would put your system under more stress, which could be a problem.

A pressure switch with the correct lower pressure limit, (cut-on pressure), should perform just as the original.

When you open a tap, the system pressure will drop below the lower pressure limit and the pump will start, and, as long as the dynamic pressure remains below the lower pressure limit, the pump will run and your flow will only be limited by the maximum delivery rate of the pump.

The over-heating motor is a question for the electrical guru's...

However, in my opinion, if a motor replacement is required, I think a new pump would be a better option. But as they say "up to you"...lol

Edited June 7, 2016 by rawhod

[Chris Pirazzi]

Some progress today.

Today I finally found all the needed parts at hardware stores and spent a few hours gluing PVC pipe (had to remove basically all the existing fittings because of the way they were glued) with a new pressure meter and a new valve nearby the pump to let water go at various flow rates for testing and pump discharging. The new valve can be turned on and off independently while totally shutting off the supply to all the bungalows. The meter was very critical for knowing what is actually going on in the pump. The nearby valve allowed me to make easy and meaningful tests and measurements without the uncertainty of whether anyone in the bungalows (which are very far away) happened to have taps on at that moment.

While gluing the PVC pipes, I also opened all the compartments of the pump that I had tools for. First thing I noticed was that the pump's built-in backflow check valve and its mysterious "Life-Extending Pressure Stablilized Unit" (which sits at the intersection of the motor output, the pressure switch, the pressure bladder cylinder, and the output) were totally full of fine, sticky mud (similar to what I found under the hopper cap before). So I spent a while getting all that mud out. I haven't been able to open the motor to look for mud on the impeller blades because I don't have the hex ratchet set needed for that.

Then I put everything together and ran the system with the test valve open just a little (to simulate one shower being on) and looked at the pressure meter. The pump was turning ON around 1.4 kg/cm² (supposed to be 2.0 kg/cm² == 28.4 PSI) and turning off lower than 2.0 kg/cm² (supposed to be 2.6 kg/cm² == 40 PSI).

Ever since the Thai "plumber" turned the pressure switch adjustment screw and declared "victory" (plus many other random people turning the screw before him), we had no idea what the pump was actually doing. Now with the new pressure gauge I was at least able to adjust the pump into near the correct range by adjusting the famous screw on the pressure switch. The closest I could get was 2.0 kg/cm² to 2.8 kg/cm² (though the pressure gauge instantly drops from 2.8 kg/cm² to 2.6 kg/cm² the instant the motor stops, so maybe it's spot-on; not sure if the "before" or "after" reading is the correct one to use).

Also with the pressure gauge we can now see that the original pressure switch (which I have not changed at any point in this whole process) is at least working in the sense of turning on and off at consistent pressures, another thing we could not determine before. Many times pump problems are because the switch turns on and off randomly because its contacts are dirty/melted, but now we know that is not the problem with our pump.

Clearing out all that mud from the insides of the pump may have helped. It seems like after I reassembled everything, the pump motor runs less often (that is, the "duty cycle" percentage of time the pump motor was running as a percentage of each complete on/off cycle seemed to go down) and the pressure bladder cylinder seems more effective at holding the water pressure up to give the pump motor a break, as it is designed to do (the needle on the pressure gauge descends slowly and evenly from 2.8 kg/cm² to 2.0 kg/cm² before the pump motor starts up again). It's hard to be sure though, because before cleaning out the mud we did not have the pressure meter or the nearby isolated test valve for good comparison. But one part I cleaned out does connect the output to the pressure bladder cylinder.

One thing is for certain: the pressure bladder cylinder from our pump did burst (which probably explains why the problem started suddenly in April, right after Songkran), and that means the screw adjustment that the Thai "plumber" made during the 15 minutes he was here would have caused the pump to burn up very quickly, because the poor motor would have kept pushing and pushing with all its might even with only one tap open partway, getting hotter and hotter until its windings' insulation melted and it shorted out. And without the "plumber's" adjustment, the pump still turned on and off way too rapidly, which also greatly diminishes its life. It was critical to replace that cylinder either way.

I tested both the new pressure bladder cylinder that my friend just sent up from Chiang Mai (which was rated at a "charged pressure" of 0.16 MPa = 1.6 bar == 23PSI, twice that of the original pump) and also the pressure bladder cylinder from the old dead pump that was in the shed (with the correct "charged pressure" of 0.08 MPa = 0.8 bar == 11.6PSI). Both of them seem to work ok.

I haven't yet had time to do a good test to see if they make the pump behave differently. A mechanical engineer friend of mine suggested that since the 23PSI cylinder has twice the nitrogen charge yet is the same size, the tank will hold half as much water when the pump reaches the cut-off pressure, and so with the 23PSI cylinder the pump should run more often because there is less of a "buffer." I will try that out when I get time.

So we might not have needed the new cylinder but it is nice to have a working spare, albeit an expensive one.

Overall, at least for now the pump is close to performing as well as it did before this whole mess started in April.

First test of course will be to see if the pump keeps performing as well in the next few days and weeks. The mud that may have played a role in this incident does not come out of the supply tanks continuously, but rather comes in massive bursts when the folks here miss their regular schedule of opening the cleaning valves at the bottom of the tanks to let out the accumulated mud, and then something happens (like a storm) to set the water in the tanks in motion. So I don't think the pump will fill with mud again in the next few days. But we will see if there is some other problem with the pump that pops up soon.

Thanks to all of you and my engineer friend and others for useful advice, and I hope this thread helps lots of future ThaiVisa folks debug their water pump!

Edited June 7, 2016 by Chris Pirazzi

[Chris Pirazzi]

Interesting: just tried a controlled experiment with the 2 cylinders. With only a low-flow shower on in both cases, I let the pump run.

With the new cylinder rated at a "charged pressure" of 0.16 MPa = 1.6 bar == 23PSI, the pump ran 7 or 8 times per minute.

With the old cylinder from the spare pump in the shed (not the one from the current pump, which is surely burst) rated at a "charged pressure" of 0.08 MPa = 0.8 bar == 11.6PSI, the pump ran 15 times per minute.

Which is exactly the opposite of what we would expect. We would expect the pump to run less (perhaps half as) often with half the PSI nitrogen cylinder charge, since there would be more (maybe twice the amount of) water in the cylinder when the pump reached its OFF pressure and thus more "buffer" for the pump.

Hmm.

Maybe the difference is there because the old cylinder is partially deflated or even has a hole in it, in which case the rated number means nothing. I did notice a fair amount of water dribble out when I removed the old cylinder, though gently poking a chopstick in the old cylinder, it clearly has a significant chunk of rubber there in place (unlike the burst one from the current pump, where the chopstick touches nothing no matter how much you shove it in there). And not as much water dribbled out of the old cylinder as dribbled out of the burst one from the current pump, though I was only using the old cylinder for about 6 hours before this test so it may not have had time to become totally water-logged yet.

So it seems it was worthwhile getting the 0.16 MPa = 1.6 bar == 23PSI cylinder, even if we are seeing double the rate of pump motor running that we could have if we had a properly specced 0.08 MPa = 0.8 bar == 11.6PSI cylinder.

[Chris Pirazzi]

Some thoughts from my friend that may help others debugging pressure tank problems:

The purpose of the pressure tank is to give the system some hysteresis.
Water is not compressible, so with no air in the system, you would have the pump turning
on/off continuously because the difference between the high pressure pump shut-off and
the low-pressure pump turn-on would be less than a cup of water (the pipes probably expand
a wee bit).

But with the pressure tank, the high pressure water can go into the tank and is stored there.
When you use enough water that the pressure in the system goes down below the turn-on
threshold, it turns on.

So... you want to MAXIMIZE the amount of water in the tank when the system pressure is
between the two pressure values. I believe you want the pressure tank to be set to just
under the low-pressure value (turn-on pressure).

If the pressure is set too low (like your first tank) then the tank is almost full at the high
pressure setting and then still kinda full at the low pressure setting. So a bunch of wasted
volume in there.

If the pressure tank is set too high, then you also have wasted volume because it doesn't
start filling the tank until you are already above the turn-on pressure.