Sizing a water system pressure pump

[sailman]

Update:  went with the Jabsco V-Flo 5.0.  It is variable speed/flow and does not need an accumulator.  Price was great at Defender.  Thanks for all the input!

 

[weightless]

With a decent PID built into the logic and some sort of hysterisis you don't need a accumulator. It's going to come on spool up slow to meet pressure and have a differential to eliminate short cycling.

Not sure it's a real problem but when you turn the tap off the PID isn't going to help much, is it? I'd expect there to be a pressure spike. I've got one of those tiny Whale surge dampers in my system. I haven't tested it without but it seems like a good idea. IIRC, it was called for in the documentation.

 

[SASSAFRASS]

^ On a high pressure setup yes it's tricky, IE thousand of PSI but on a positive displacement pump at low pressure no, you shouldn't see a significant spike.  A good logic loop or PID will spool up and down as the target is required and approached over a time factor.

A practical example would be everything open shower sink etc then one or two off the pump will slow down ahead of the setpoint as the rate of rise increases even though the set point has not been achieved.  The higher the operating pressure the harder it is to remove surges but at lower pressure it's much easier.  Have setup direct drive piston pump hp washdown units with multiple pumps where you are turning very slow revolutions, single value rpm and then yes it takes alot of tuning.  Something like the Marco pump probably doesn't have a crazy amount of logic but with variable speed will most like eliminate and surges. Most pumps used are up down pressure switches with only a differential so a accumulator will help short cycling, but as above with logic in the control it's not needed.

 

[blurocketsmate]

@climenutsis right, a modern, constant-pressure pump sans accumulator is the only way to go.

 

[DDW]

My experience with normal pumps w/accumulators, Jabsco variable speed, and Marco variable speed (currently have all those setups in one thing or another).  Normal pump with accumulator works, but there is pressure variation between the cut in and cut out point, and the accumulator does not smooth out the pulsing of these multichamber pump designs. The Jabsco ones work - for awhile - and keep pressure pretty constant. There is a slight drop in pressure and lag until it come back up again when you turn on a tap (maybe a half second total time). If there is a pressure spike when you turn off the tap it isn't apparent. There are still pulsations due to the multi chamber design. I've had 3 of these fail over the years. The Marco pump is a very different design, a gear pump pushing through a check valve. No pulsations and a very different sound. The pressure is higher than a typical Jabsco or Whale, even with a smallish one (the 3 is what I have). They are not self priming, can be primed by cracking the priming screw up to about 4-5 ft head. But if there is ANY suction leak on the suction side you will have problems. For example the supplied filter had suction leaks. Their control circuit is very good, better than the Jabsco at keeping constant pressure. The lag when turning on a tap is short, but seems to get longer if the suction head is more. E.g. it gets longer as the tank level drops. The pump is noticeably better constructed than a Jabsco or Whale or Sureflo. It has BSPP threads on the in and out (they supply barbs but if you want to plumb it...).

I'm pretty happy with the Marco in the RV, replacing a Jabsco variable that went intermittent. But it has to suck about 2 feet and this causes occasional problems. I've bought one to put in the motorboat (replacing a Whale and accumulator) since the pump is below the tank and that solves all of those issues. I may not replace the one in the sailboat as the pump is about level with the tanks and I am changing tanks all the time when one runs out - that might lead to having to prime the Marco each time which would be a PITA. 

Both the newer Jabsco variables and the Marco have diagnostics and safety features, like they will time out on a broken line and also detect a slow leak. There are many complaints online about the Jabsco reporting errors falsely and shutting down, haven't seen much negative on the Marco.

 

[IStream]

So I pulled the accumulator out and there's no discernable difference in performance. That gave me room to install a water filter and everything is working great. 

 

[DDW]

IStream, how much delay do you get when turning on a tap? And where is your pump in relation to the tank?

This is only annoying to me when taking a shower, as I have the trigger shower head. Pull the trigger and there is a momentary droop in pressure. It has varied a bit and seem to do with any air in the suction side, and also the water level in the tank. I have had trouble with suction leaks on that side. 

 

[SASSAFRASS]

If you have a decent strainer in the system, you can remove the sediment screen in the shower head, usually makes a difference by increase in flow. If you have a household mixing valve same thing can really restrict initial flow. Sure you already know but line size on these is huge, going up on suction and discharge can be a big help.

 

[IStream]

IStream, how much delay do you get when turning on a tap? And where is your pump in relation to the tank?

This is only annoying to me when taking a shower, as I have the trigger shower head. Pull the trigger and there is a momentary droop in pressure. It has varied a bit and seem to do with any air in the suction side, and also the water level in the tank. I have had trouble with suction leaks on that side. 

TLDR version: I've got no pressure drop issues but it's probably because even without a dedicated accumulator, there's so much compliance in my water lines that I'm effectively still running with an accumulator. Before removing the dedicated accumulator, I found that priming usually worked but not always. I need to re-test this now that the accumulator is out and think it'll be more reliable.

So, here's the whole picture. My boat was designed for charter, so it has four cabins and 6 water tanks (no typo), totaling 440 gallons. All the tanks are all glassed in and epoxy-lined sections of the hull running fore and aft and flanking the bilge, with the top of each tank just below the floorboards. Each tank has a supply line that runs to a manifold with 6 inputs, each input with a ball valve, and with one output that leads to the pump about six horizontal feet away. The pump is mounted at floorboard level so the suction head ranges from about 1 foot to 3 feet, depending on the water level in the selected tank.

When I first put the Marco/Mate pump in I had priming issues that I ultimately traced to very tiny air leaks in the PO's input distribution manifold, which consisted of a cobbled together collection of gray CPVC plastic tees and valves. It turns out that the CPVC valves he used were only rated for pressure and not vacuum but they were good enough to last for 10 years or so while he owned it. Once I figured this out I  replaced the manifold with a monolithic aluminum block and decent ball valves and that solved most of the priming problems. That said, before removing the accumulator I found that if I ran the supply lines dry and didn't crack the bleed valve, the pump might hit its 10 second priming timeout before it gets the lines filled. In that case, I'd cycle power to the pump and try again. Worst case I would open the galley faucet, which has a sprayer hose, and suck on the output to help get the air out. The accumulator could drive decent flow for about 10 seconds so I think it was contributing to the timeout issue.

On the output side, there's hot and cold clear hose running to all four heads, the hot water heater, and the bow washdown spigot, so there's probably well over 100 feet of compliant tubing that acts as an accumulator. After I pulled the dedicated accumulator out, I pressurized the system and turned off the pump and found that it would drive decent flow for about 2 seconds. That seems to be enough to prevent the issue you're seeing in your shower even at my most distant faucets and I'm hoping it's a sweet spot that'll let the pump prime within its timeout window without heroics, but I still need to test that. 

FWIW.

 

[DDW]

Well you have confirmed by suspicion that the suction side has to be truly vacuum tight. Even the typical Whale shark-bite type quick connects aren't vacuum rated. I think the accumulator probably contributes to the priming issues because it can develop very little pressure when dry, and has to push that through the internal check valve. If it also has to do that against pressure in the accumulator it can't make matters easier. I cannot get it to prime into a shut system, I have to open a tap while I prime. Once the gears get a little wet it takes off. A solution I've thought about is to put a barb fitting in place of the priming screw, tube on it running to a convenient location where a valve could be cracked. The automated version of this is a tube or fitting with a restrictor allowing only a tiny bit of water through (but a lot of air) that returns to a tank. Since the pressure sense is on the other side of the check valve, seems to me this would work and make is essentially self priming. Try it and report back - save me the trouble . 

All of this makes it sound like the Marco pump is a pain the butt, actually most of the time it works just ducky. 

 

[El Borracho]

Plumbing a water system to function with suction is asking for trouble. Here at the cabin and on the boat plumbing for gravity priming of pumps makes a world of difference regardless of what the overly optimistic pump  datasheets say. Sometime one has no choice....

Air is about 1000 times thinner than water. A less than pinhole leak in a fitting or pump flapper valve makes for unhappy sailors.

 

[IStream]

Well you have confirmed by suspicion that the suction side has to be truly vacuum tight. Even the typical Whale shark-bite type quick connects aren't vacuum rated. I think the accumulator probably contributes to the priming issues because it can develop very little pressure when dry, and has to push that through the internal check valve. If it also has to do that against pressure in the accumulator it can't make matters easier. I cannot get it to prime into a shut system, I have to open a tap while I prime. Once the gears get a little wet it takes off. A solution I've thought about is to put a barb fitting in place of the priming screw, tube on it running to a convenient location where a valve could be cracked. The automated version of this is a tube or fitting with a restrictor allowing only a tiny bit of water through (but a lot of air) that returns to a tank. Since the pressure sense is on the other side of the check valve, seems to me this would work and make is essentially self priming. Try it and report back - save me the trouble . 

All of this makes it sound like the Marco pump is a pain the butt, actually most of the time it works just ducky. 

I put a Sea Tech / Watts quick connect fitting on the input to my pump to facilitate servicing and it seems to be airtight in suction, at least for now. Everything upstream from there to the tanks is barbs and clamps. @El Borracho I agree that it would be great to have a gravity prime but it's just not practical with my tank arrangement. 

When priming, I always open a tap to atmosphere. It's easy to do and stacks the deck in your favor. As you said, it's a good pump and mine has never lost prime by itself. It only happens if I run a tank dry and let the pump run long enough after that to blow most of the water out of the gears. Even then, when I switch to another tank it'll usually catch a prime with one or two tries. On the odd occasion it doesn't, my suck on the galley faucet trick does the job reliably. 

Just for you, I'll do various priming torture tests next time I'm at the boat but I don't plan to crack the priming screw if I don't have to. I just got the galley cabinet where the pump is mounted all mopped out and dry following all my misadventures getting the accumulator out and the new in-line filter in. Pro-tip: check that the store gave you the 1/2" PEX pipe you asked for and not the 15mm stuff before you put everything together and turn on the pump. That UP12/E can move a lot of water in a hurry.

 

[DDW]

On the sailboat I've got the Whale fittings everywhere, and there are a few between the tanks and the 4 way selector valve feeding the pump. Even with the Jabsco there have been problems, particularly with one tank. I chase around and and wiggle or disconnect-reconnect the fittings and get it to work. On the RV it's those cheap compression-sort-of fittings but a pretty short run. Helps to have a nice clear tube going into the pump, you can see if there are bubbles. Suction side leaks are insidious. When first built I could not get more than about 15 gallons of diesel out of one tank before it would suck air. I even vacuum tested it. We finally took the whole run apart, found nothing, put it back together and it has worked perfectly ever since. 

 

[SASSAFRASS]

Not always, but sometimes a suction riser in the line before the pump can help nuisance priming issues when you can't get head on the pump suction. Using the self vulcanizing rescue tape on tiny vacuum leaks in fittings will help too.  Vacuum leaks are a major pain, used to run a ancient 200ton three stage evaporator water maker.  Used the water reacting fiberglass tape on fittings all the time when chasing leaks. All sweated cuni. Unions.