Cavitation issues

[Bob Rohr]

Those last 10- 20 degrees over 170 are the toughest to get exchanged, as you are learning This is true with closed loop also. As the delta T shrinks the exchange slows.
It is doable, but with the deck stacked against you like that it will be harder to accomplish. The conditions you are working with, high head circulator, open system, large pressure drop circuit are adding up, working against you.

Temperature and pressure, those are the two factors involved in the cavitation issues I suspect you are experienced. Change one or both and you chances for success go up.

With an open system the PONPC, point of no pressure change is at the actual water level, so pumping into the boiler is not a viable solution, pumping away adds the circulator ∆P to the circuit, which you dearly need.

 

[maple1]

This has been a very interesting thread - Catch-22's all over the place.

 

[Marshy]

The 20' of head loss you came up with seems high to me.

1" pex has a pressure loss of 11.4 psi per 100 ft, or 1.14 psi per 10 ft at 15 gpm. If you convert that to feet of head, 1.14 x 2.3 = 2.622 ft of head loss for 10' of pex. So you have about 10.5' of head loss in 40' of pex if you pumping 15 gpm, and less loss as flow goes down.

Actually, 0.55 psi head loss at 10 gpm. 0.55 x 2.3 = 1.265' per 10', or 5.06' for 40' of pex.

What head loss are you assuming for the htx?

I see four 45 elbows on the copper side and a zone valve. I also see a strainer at the pump discharge and a 90 elbow. Then returning into to the boiler you have a 45, the diverter valve and a 90 into the boiler jacket. Maybe 20' is reasonable?

One thing you absolutely cannot have is a pump with a discharge head less than the system resistance (head loss). You won't move any water.

 

[warno]

@Marshy here's a list of every fitting in my boiler loop.

BV = full port ball valve.
Everything is 1" unless otherwise noted
The diverter has a CV of 7.5
The themomix has a CV of 10
A similar FPHX on the site has a pressure drop off 2.7PSI on boiler side.

 

[warno]

I'm looking for a new FPHX and I'm wondering if there is a noticeable difference between 1" port and 1.25" ports? If I go with 1" MPT ports I can reuse my exsisting parts.

 

[Bob Rohr]

If the boiler connections are 1" and piping are 1" that will be one of the limits on how many BTUs you will get out of the beast.

What about combining the boiler connections with reverse return, upsize to 1-1/4" and feed directly to the HX in the boiler room. That gives you the best chance of getting full output, at least to the HX, and could probably run with a smaller, flat curve circulator. So the cavitation issue goes away.

Yes you may still need the large pump to overcome the small tube and pressure drop into the building, if in fact you want 12- 15 gpm.

I'd abandon the 190° target SWT at the heat emitters. 180 should be plenty for any hydronic AH, fin tube, etc. even 160. That would also eliminate trying to drive the boiler so high and always be at that fine line of boiling over. Probably get away from any dump zone if you don't drive it so hot.

Really the only reason to run that 190 would be to store in a buffer and pull off it down to say 100 for radiant or low temperature emitters, then you have a huge 90∆ to work with.

The trend in the industry is to design hydronic systems to cover the load on design days with no more than 120F! That is a commonly used design temperature in Europe with radiant floors, ceiling and walls or generously sized panel rads. With a properly sized coil an AH could run down around 140 SWT.

Piping losses of course go down with lower operating temperatures also, and wood consumption.

Lets say your wood to water efficiency is 70%. The higher the operating temperature the greater all the losses and efficiency hits, all along the line.

Heat loss and heat transfer is all about ∆T. The greater the delta, the higher the heat exchange, including loss from the boiler jacket, piping, HX, underground, etc.

 

[warno]

@Bob Rohr The one and only problem with putting the HX in the boiler shed is excess heat dump. With running the boiler lines into the garage first then I can dump any over heat situation into a building that I actually want heated. I know the dump zone won't run all the time or with every burn, but if I do get into a over fire situation I'd rather use that heat to my advantage. I had actually thought about that during this discussion in this thread.


I just a had thought and I want to ask a question before presenting my thought.

My cavitation issue on my boiler pump is because I have a high head loss number on the discharge side vs a low pressure number on the suction side, is this correct?

 

[Marshy]

...
My cavitation issue on my boiler pump is because I have a high head loss number on the discharge side vs a low pressure number on the suction side, is this correct?

No, your cavitation issue is strictly related to the pressure and temperature of the fluid entering the pump. It has little to do with the head loss on the discharge side of the pump. You either need lower temp water entering the pump or higher pressure. Only way to increase pressure is to lower the pump in relation to the water level in the boiler jacket. You can also reduce the head loss in the suction piping but you have so little to begin with its unlikely to make a difference. You also have so little extra room to lower the pump it may not cure it either. That is why lowering the water temp may be the ultimate cure. That or try a low head circ pump because it does not need as high a pressure at the inlet of the pump (required NPSH) which means you can run a hotter water temp. Clear as mud?

Just for clarity, the system restrictions on the discharge (head losses) do affect where the pump operates on its performance curve. Depending on the pumps characteristics the required NPSH curve could be fairly flat, however the NPSH required usually increases as the discharge flow from the pump goes up. Therefore, adding restriction to the pump discharge lowers the potential for cavitation, albeit minor.

This picture is an example of the pumps operating curve and NPSH required. It also shows the system resistance curve. If you add more resistance the pump puts out less flow and more pressure, you move back (left) on the operating curve. The NPSH required for that operating point can be determined by drawing a straight line down to the NPSH curve. If the curve is flat adding restriction does very little good.

 

[warno]

Well my thought was plumbing in this configuration. The first circ coming off the boiler would be as low to the ground as possible and all suction side piping would be 1.5".

But now that you say the head loss isn't that important I don't think this would matter. My thought was the first circ would add pressure to the circ pumping into the garage. But with what you just said it makes me think the first circ, in this configuration, would cavitate just the same.

 

[Bob Rohr]

Here are some formulas to show how and when cavitation is possible. I think you have 1" piping so that factor in the example is correct, not sure about you exact flow rate, assume one and try a few others.

It sounds like you have already seen or heard cavitation when you change pump speeds so you need to make some changes.

Several options have been presented, if you run the math it will predict which will or won't work.

I hate to see you change piping and locations if they will not solve your issues.

I'd be glad to send you a Quicksetter so you can observe some actual flow rates if that helps. Just know a 1" Quicksetter only goes to 10 gpm. Installing an 1-1/4 on 1" piping doesn't make a lot of sense as your piping is undersized for the higher flow rates, that is why I suggested combining the two 1" boiler connections so you have the ability to handle the flow rates I think you are trying to target.

 

[jwise87]

Here are some formulas to show how and when cavitation is possible. I think you have 1" piping so that factor in the example is correct, not sure about you exact flow rate, assume one and try a few others.

It sounds like you have already seen or heard cavitation when you change pump speeds so you need to make some changes.

Several options have been presented, if you run the math it will predict which will or won't work.

I hate to see you change piping and locations if they will not solve your issues.

I'd be glad to send you a Quicksetter so you can observe some actual flow rates if that helps. Just know a 1" Quicksetter only goes to 10 gpm. Installing an 1-1/4 on 1" piping doesn't make a lot of sense as your piping is undersized for the higher flow rates, that is why I suggested combining the two 1" boiler connections so you have the ability to handle the flow rates I think you are trying to target.

A down and dirty way he could check flow would be to pull the return off at the boiler and dump into a bucket. Obviously not at operating temps. That would give an idea of the flow to punch into the formulas.

 

[warno]

@Bob Rohr I'm a little confused by when you say adding 1 1/4 to 1" piping? And by pulling from both 1" connections do you mean pull from both my supply ports and return into both my return ports?

 

[Marshy]

@Bob Rohr I'm a little confused by when you say adding 1 1/4 to 1" piping? And by pulling from both 1" connections do you mean pull from both my supply ports and return into both my return ports?

I think he was saying to connect the lower boiler port to the suction of the pump (combine the upper and lower) in an effort to reduce head loss on that side of the pump. That's how I understood it.

While it would reduce some head loss, my only hesitation about that is if you still have 1" pipe from where the two tee together up to the suction of the pump you haven't taken full advantage of eliminating the restriction. Converging flow in a tee is fairly restrictive and if you are still using 1" after that tee I don't think its worth the effort. That is why I suggest just going with 1.5" pipe on the whole suction side. On second though, you could connect both ports together as suggested but use 1.5" pipe. I would add full port ball valves to each port and run it just drawing from the upper port. If its cavitating open the lower port and see if it goes away. Also try just the lower port as the colder temp would also be beneficial. Use a long sweep elbow where ever you can on the suction side. (I feel like that was a small tangent, sorry.)

Will either option solve the problem? I don't know without making some calculations (and assumptions since we don't know the flow rate). Good advice was given about how to determine the flow. If you get the chance try to use the 5 gal bucket method and estimate flow.

I could attempt to run some numbers for you but I would need some dimensions from you. From the floor of the boiler cabinet what is the distance to the water level in the boiler jacket? Can you total up the straight copper tubing leading up to the HTX and also back to pex return?

 

[Bob Rohr]

yes increase to 1-1/4" when you combine upper and lowers. Think of it as building a manifold where a bunch of small ports enter into the larger trunk piping. This allows you to flow more through the boiler and HX. Install the HX in the building with the boiler. With a properly sized HX you should be abl;e to get most all the energy out of the boiler and into the box (HX).


With a non gasification boiler I/ you might be tempted to remove the return mixing valve. Rarely do OWF designs worry about return temperature, especially at elevated temperatures that you want to run. Only on cold start up would you have some condensation.

Or put the return temperature protection on the B side. We are trying to simplify the A side piping and pressure drop as much as possible.

Our goal is to get as much heat out of the boiler, reduce cavitation, and be able to still run high temperatures. At least I think that is the goal?

Pump performance could be measured by measuring pressure on inlet and outlet, refer to the pump curve. pretty easy to get a few 1/4 gauge ports into the piping. I've seen guys use the cheap saddle valves to get into a pipe for pressure readings if you want quick and simple.

 

[Bob Rohr]

P/T ports aka pressure temperature. These small fittings are often included throughout a piping system. Officially they are named after the inventor Peterson. They allow you to "needle in" a gauge to read pressure or temperature. Mechanical engineers use them often on systems to check performance and trouble shoot.

Think of a basket ball inflation connection, that is how PT ports work. Spit on the needle and slide it in. Available all the way down to 1/8 connection side, easily tapped into piping after the fact. Larger pumps often have threaded openings to add PT ports.

 

[warno]

@Bob Rohr or @Marshy I'm trying to do that math what would I enter for my psi entering the circ since its an open system? Wouldn't that be the weight of the water pushing down? If so, isn't that the .435 psi/ft we all talked about earlier?

 

[warno]

based on 10 gpm and 190° water are these my numbers?

 

[Marshy]

@Bob Rohr or @Marshy I'm trying to do that math what would I enter for my psi entering the circ since its an open system? Wouldn't that be the weight of the water pushing down? If so, isn't that the .435 psi/ft we all talked about earlier?

The short answer is yes, the long answer is no.

Long answer: No, the NPSH available is the pressure of the water column minus the head loss from the pipe and any fittings.

Check this NPSH calculator out.

Adjust values to accommodate you actual setup. I guessed the flow rate (10 gpm). All other parameters should not be a guess. The piping material is cast iron unlined.

http://inventory.powerzone.com/resources/npsh-calculator/:al=600:alu=Feet Above Sea Level:fs=1:sp=48:spu=Inches:fr=10:fru=GPM:vc=Water:ft=194 Deg~p F (90C):v=0~p000315555:vu=KGM-1S-1 / NSM-2:sg=0~p965:vp=10~p17:vpu=PSIl=2lu=Feetid=1idu=Inches:hzc=Cast-Iron - new unlined:hzfc=130:rh=0~p011811023622047244:rhu=Inches:rfvu=FT/Sec:rnu=PSI:fits=Elbow, 90° SR|1!Valve, Ball, Full Port|1!

 

[Marshy]

Here is another good reference for manual calculations.
https://www.engineeringtoolbox.com/npsh-net-positive-suction-head-d_634.html

 

[warno]

Here's what i came up with for cast iron 1" @ 10gpm. Is this right? Good or bad?

 

[Marshy]

View attachment 221037

Here's what i came up with for cast iron 1" @ 10gpm. Is this right? Good or bad?

A negative NPSHA would mean its cavitating before it makes it to the pump lol. You must of mess something up. If you could post pics of the rest of the inputs I will help you.

Piping system section is for the section leading up to the pump only.

 

[warno]

Piping system section is for the section leading up to the pump only.

Well this might help a bunch. Lol let me run it again.

 

[warno]

here it is again @Marshy

 

[Marshy]

here it is again @Marshy

View attachment 221041

View attachment 221040

View attachment 221039

You still have one possible mistake that I can see. If you measured from the centerline of your pump to the top of the water level in the boiler jacket, is it only 9 inches in distance?

 

[Marshy]

@warno, to also have four 45 elbows, not two. Don't forget about the two you used to make the turn out of the boiler jacket.