Cavitation issues

[BoiledOver]

After thinking through my post, I realize I stated lowering the circ would decrease pressure. Quite the opposite. You are looking to increase outlet pressure to cure your cavitation. But, a reduced boiler loop, larger FPHX, and different circulator may address the cavitation issue anyway.

 

[Marshy]

If I just pull straight off the water surface it will draw air into the pump. The little drop down piece is a dip tube to insure the pump doesn't suck air in.

Will the added height of the pipe headed to the pump benefit any from doing this? I mean As far as the water column pressure goes.

No.
The pumps I drew all have one thing in common, the same amount of static head (ignoring pipe friction loss). That's because the pump centerline to the water level is the same.

 

[warno]

The circulators I'm running right now are a B&G NRF-36 and a Grundfos 15-58. With this new drawing should I run the 15-58 at the boiler side and the NRF-36 at the storage side?

here's my new drawing I just scribbled up.

 

[warno]

@Marshy Thank you for the drawing. That makes sense. I did finally realize that my thought on ther idea was irrelevant because the water level is the same pushing down regardless where my port is in the water jacket.

We've began kicking around piping ideas now. I think we are getting somewhere.

 

[BoiledOver]

@Marshy Thank you for the drawing. That makes sense. I did finally realize that my thought on ther idea was irrelevant because the water level is the same pushing down regardless where my port is in the water jacket.

We've began kicking around piping ideas now. I think we are getting somewhere.

Pertaining to the garage side: The B&G might be just right on the storage loop. However your location doesn't seem to sit right, shouldn't it be pulling from the diverting valve with the diverting valve plumbed on the cold side? Just a tee on the hot side.

Pertaining to the boiler room: Should the circ be pulling from cold side of the FPHX? 4 Isolation valves at the FPHX for maintenance? Or do they have a feature built in?

Maybe 2 15-58's would do the trick.

 

[maple1]

After looking at your last sketch I am thinking more that your dumping circuit could be redundant.

In an overheat situation - the diverter would direct flow around the dump zone HX, rather than through storage. But, unless storage is getting close to charged, storage usually would be able to send colder water back to the boiler, than the dump HX would with its dT, therefore have more potential to cool the hot boiler better. My storage stays quite cold on the bottom, until first lap of charging is done. Then it finishes up rather quick with second lap. But, then again, maybe it is better if it did happen when storage is getting topped - or if you have a lot of mixing when charging storage.

That might be more rambling to be spit aside with other stuff that has since been spit aside....

I will second BoiledOvers comment about the storage circ being not quite in the right place - it should be downstream from the T on that return line. The boiler circ comment - hmm, back to cooler inlet temps vs. increased inlet pressure drop.

 

[warno]

After looking at your last sketch I am thinking more that your dumping circuit could be redundant.

In an overheat situation - the diverter would direct flow around the dump zone HX, rather than through storage. But, unless storage is getting close to charged, storage usually would be able to send colder water back to the boiler, than the dump HX would with its dT, therefore have more potential to cool the hot boiler better. My storage stays quite cold on the bottom, until first lap of charging is done. Then it finishes up rather quick with second lap. But, then again, maybe it is better if it did happen when storage is getting topped - or if you have a lot of mixing when charging storage.

That might be more rambling to be spit aside with other stuff that has since been spit aside....

I will second BoiledOvers comment about the storage circ being not quite in the right place - it should be downstream from the T on that return line. The boiler circ comment - hmm, back to cooler inlet temps vs. increased inlet pressure drop.

The heat dump is defiantly for when storage is charged and still have a fire burning.

I agree with the storage circ being wrong. I redrew that part.

I was wondering about the adding inlet drop if the boiler circ was on the return side. I think it might be alittle to much.

@BoiledOver with strainers on both sides of the FPHX there shouldn't be much gunk in there. If it does plug up I have webstone propal valves at the ends of each line. I would just blow the lines out and remove the HX via unions.

 

[BoiledOver]

@BoiledOver with strainers on both sides of the FPHX there shouldn't be much gunk in there. If it does plug up I have webstone propal valves at the ends of each line. I would just blow the lines out and remove the HX via unions.

Gotcha, you already had your plan for that.

Maybe @Bob Rohr or @Marshy will review the direction you are currently considering. Attached is a thought on how I would approach the design at the boiler. Enlarged discharge and circ down low, the return temp should rarely exceed 170F.

I have a hangup about keeping equipment (circs) cooler rather than warmer, in most cases, hence on the return. Just for a hoot, consider swapping the circs to see what happens before making any mods. Ya might get some feedback from the results.

 

[warno]

I guess if I did put the circ on the return side of the HX I could pull from the bottom of the water jacket since that is where the hottest water is.

I could over size the piping going to the HX and up to the pump. The pump itself would still be below the water level by about 48" in that location and the only major head loss would be the HX and the themomix valve.

What does everyone else think about that idea? Too much head loss for the suction side?

 

[maple1]

With that big hx you were looking at last I don't think that would be an issue. Looked like hardly any pressure drop I think?

 

[BoiledOver]

I take it your boiler is not a sealed type you can maintain any pressure in?
If so, you're going to find 180-185* is the maximum temperature you can run without destroying pumps.
Also keep in mind that the higher the "head" is in the piping the more NPSH you need to maintain on the inlet of the circ.
Larger diameter tube on the discharge side is the only thing that will accomplish that all other factors being equal.

#01-From @heaterman Is this correct?

This. You need to lower the pump and reduce the temp of the inlet. Enough of either one or a mix of both will solve the issue.

#02-From @Marshy Is this correct?

I think the easiest thing I could do is replumb the suction side with bigger piping and move the circ to the floor level in a horizontal position.

#03-From you. This is contrary to #01. Whichever is correct can be implemented.

A bit of a sidebar, the noise you are hearing and calling cavitation could alternatively be the noise of radial thrust. Still an issue but seems to be a different symptom of the same contributing factors. (broken link removed to http://blogs.heattransfersales.com/blog/hvac-pumps-radial-thrust)

If #01 and #02 are correct, you can leave the FPHX in place but change your plumbing as the image shows. You showed a similar piping scheme in your post #61 on page three although it was pulling from the boiler (hot) rather than pulling from the FPHX (cooler).

 

[warno]

@BoiledOver the only thing wrong with leaving the FPHX in the garage and pulling from the return side was the tremendous head loss to the suction side of the pump. I don't remember what post it was but we've discussed it here somewhere. Lol

I'm almost 100% certain the noise from the pump is caviation. It sounds like pebbles are running through the impeller. From what I've read that is the sound of the air bubbles bursting.

 

[warno]

I drew up some scenarios of piping just the boiler side with the circ pulling from the return side of the HX.

The hottest area of the boiler is the lower left corner and the coolest is the upper right corner.

The larger looking pipes will be 1.25" and the smaller will be 1".

Which do you guys think will work best.

 

[maple1]

I would do 1. But reasoning is sort of convoluted.

I would expect the hotter spot to be at one of the top ports. And cooler to be one of the bottom. And both top to be hotter than both bottom. But it doesn't seem to be that way - which might be a sign that when you are burning & circulating water thru the boiler, the flows through it aren't even and there are places inside that the water could be moving faster through. Or others where it is stalling. So if you go 1, that will make the biggest pipe at the places of biggest expected flow (hot top left & cool bottom right - where you are piped in now), but with ball valves at all outlets still let you bring flow to/from the other two ports to help you balance the flow through the boiler. I would likely have said connect the two top ports to supply, and two bottom ports to the return (call that 4?) - but what you said about the top right port is coolest has me a bit boggled there. Internal flow must be channeling up the right side, then over to the left along the top? But I would still expect to see bottom right (where you are returning to now) the coolest. Are those relative temps measured at the ports? Or are there front temps thrown in there too? I could see how bottom left might be hottest if not much water moving inside at that spot but can't see how top right could be coolest if return is coming in bottom right.

Depending how hard it is for you to do, I might do it all in 1-1/4" also. Then balance with ball valves.

 

[warno]

@maple1 the temps I measured the hottest were on the water jacket bottom plate. The bottom plate just under the lower left port was reading right around 200° when the upper port itself, same side, was reading closer to 190°. The lower left side of the bottom plate was just a few degrees less then the right. But the upper left port, very little to no flow there, was reading 170°. Right now the lowest flow points are the upper left and lower right. I'm looking to alleviate that by trying to get a cross flow affect. Because my firebox is only submerged by water on it's upper half it kind of makes sense the lower water jacket would be hottest. I also get a hot spot right at the front plate, by the wood loading door, of the water jacket because little flow gets up there.

Here's a pic you can see my water jacket bottom plate compaired to the firebox body. The pipe sticking in there is my current return port.

 

[BoiledOver]

@maple1 the temps I measured the hottest were on the water jacket bottom plate. The bottom plate just under the lower left port was reading right around 200° when the upper port itself, same side, was reading closer to 190°. The lower left side of the bottom plate was just a few degrees less then the right. But the upper left port, very little to no flow there, was reading 170°. Right now the lowest flow points are the upper left and lower right. I'm looking to alleviate that by trying to get a cross flow affect. Because my firebox is only submerged by water on it's upper half it kind of makes sense the lower water jacket would be hottest. I also get a hot spot right at the front plate, by the wood loading door, of the water jacket because little flow gets up there.

Here's a pic you can see my water jacket bottom plate compaired to the firebox body. The pipe sticking in there is my current return port.

View attachment 221597

Just so ya know, take my advice with salt. I am just a fella who likes the DYI approach too. I had a pretty big (not dangerous) fail prior to getting an Eko boiler. Failed the research and engineering part.

That is a rather unique design you have built from. Is it all yours? Now I see why you have the oddball temps at the ports. You sure do have the welding skills, kudos to ya.

To me, supply from the top and return below is logical. So, I would do #3 but reversed. Combining both lower ports and both upper ports should help in keeping the top hotter than the bottom, instead of a criss-cross or side-to-side fashion.

If your plan includes a larger FPHX that has less head loss and you move it to the boiler shed, you should see big gains in the boiler loop efficiency. The results may put a damper on the new storage loop. If your storage loop is pressurized, no problem. If on the other hand it is unpressurized will you be able to get the flows needed to take all of what the PFHX is providing at 200F? Without the cavitation issue moving to that side? Maybe your original HX will keep up if you throttle back on the boiler output. I know you stated you wanted the big numbers in storage.

190-200F water temps in the boiler at regular operation, holy cripes. Is that how the owb fellas go? Those temps in storage will be sure to keep your garage warm, lol. The higher the storage temps means higher standby losses. During the coldest spells, I will run storage temps up to 190-192F. Generally 180-183F is the target.

After researching, I see where the larger pipe diameter is suggested on the inlet side to help in reducing cavitation issues. Anyway, it looks like you are on the right track with your plan. Circ low and away from the HX with larger piping.

 

[warno]

Just so ya know, take my advice with salt. I am just a fella who likes the DYI approach too. I had a pretty big (not dangerous) fail prior to getting an Eko boiler. Failed the research and engineering part.

That is a rather unique design you have built from. Is it all yours? Now I see why you have the oddball temps at the ports. You sure do have the welding skills, kudos to ya.

To me, supply from the top and return below is logical. So, I would do #3 but reversed. Combining both lower ports and both upper ports should help in keeping the top hotter than the bottom, instead of a criss-cross or side-to-side fashion.

If your plan includes a larger FPHX that has less head loss and you move it to the boiler shed, you should see big gains in the boiler loop efficiency. The results may put a damper on the new storage loop. If your storage loop is pressurized, no problem. If on the other hand it is unpressurized will you be able to get the flows needed to take all of what the PFHX is providing at 200F? Without the cavitation issue moving to that side? Maybe your original HX will keep up if you throttle back on the boiler output. I know you stated you wanted the big numbers in storage.

190-200F water temps in the boiler at regular operation, holy cripes. Is that how the owb fellas go? Those temps in storage will be sure to keep your garage warm, lol. The higher the storage temps means higher standby losses. During the coldest spells, I will run storage temps up to 190-192F. Generally 180-183F is the target.

After researching, I see where the larger pipe diameter is suggested on the inlet side to help in reducing cavitation issues. Anyway, it looks like you are on the right track with your plan. Circ low and away from the HX with larger piping.

My design came from a manufactured boiler that is on the market. I changed some things to better match the parts I had to use. Thanks for the compliment.

The storage is pressurized so the storage pump will like that. If I throttle the heat transfer any lower with my current HX my boiler goes into idle mode.

I'm not meaning to run 190° water all the time but once my storage hits 170° the heat transfer slows to a crawl and the boiler tends to get really hot really fast. With forced combustion I'm kind of stuck with the firing rate my fan drives air into the box. I am also getting secondary fire now so that means more heat. I'm planning on switching to induced draft and being able to slow my firing rate substantially. If I can hit 180s in my storage with ease I'd be happy. After reading the temps everyone else uses that should be good.

 

[Bob Rohr]

I'd agree, combine the boiler connections with reverse return piping, increase to 1-1/4 at the second connection.

The pump as low as possible.

The HX at the boiler.

Eliminate as many "devices as possible in the boiler piping. I'd use a dirt separator not a Y strainer. A brand new Y strainer has a few feet pressure drop, much worse as they plug up. Near zero pressure drop in a separator.

Any time the boiler pump is off the boiler will stratify, basic thermodynamics, I'd use the top ports as supply connections.

Put the dump device, if you really need it, on the B side, that anti condensation valve, better in the pressurized side of the HX as the valve also has some pressure drop.

Long sweep turns are much better, less pressure drop.

I suspect with these changes you could use a low head, flat curve circ and help lessen cavitation potential. Something like a Taco 0010 might get that 12- 14 gpm you are after.

You will still need a high head circulator on the B side to move 10GPM or more. You can put pumps in series instead of paying the high $$ for high head circulators.

 

[warno]

I'd agree, combine the boiler connections with reverse return piping, increase to 1-1/4 at the second connection.

The pump as low as possible.

The HX at the boiler.

Eliminate as many "devices as possible in the boiler piping. I'd use a dirt separator not a Y strainer. A brand new Y strainer has a few feet pressure drop, much worse as they plug up. Near zero pressure drop in a separator.

Any time the boiler pump is off the boiler will stratify, basic thermodynamics, I'd use the top ports as supply connections.

Put the dump device, if you really need it, on the B side, that anti condensation valve, better in the pressurized side of the HX as the valve also has some pressure drop.

Long sweep turns are much better, less pressure drop.

I suspect with these changes you could use a low head, flat curve circ and help lessen cavitation potential. Something like a Taco 0010 might get that 12- 14 gpm you are after.

You will still need a high head circulator on the B side to move 10GPM or more. You can put pumps in series instead of paying the high $$ for high head circulators.

I'll look into the dirt separator again. It was a matter of cost last time is why I went strainer. Might still be an issue. Not sure yet.

@Bob Rohr we've been talking about running the circ on the return side after the HX to help eliminate the cavitation issue. Would you agree with this or no? It looks like your drawing has the circ on the supply side.

Another question being the boiler protection valve I believe has to go on the boiler, open, side of the HX because the boiler water has to directly affect the thermostat inside the valve. Is this correct thinking? I'm just not sure how it would work on the other side without the heated water to open it.

As far as pumps go, I'm hoping to use my current 15-58 and NRF-36 I have now. At least to see if they can do the job before I buy another pump/s.


For separators how is this one?

 

[warno]

I did find a hiccup with the return protection valve. It's a Caleffi Themomix valve with 1" MPT ports. The bigger option for the 1.25" valve is another $200 i would have to spend. So I think I'm going to run my 1" valve for now and if I see issues with it cutting flow to much I'll address it with a larger option. I'll have all the plumbing be 1.25" everywhere else but the valve. I do know this will be the bottle neck, but the 1.25" themomix valve has the same CV of 10 as my 1" valve. so other then the restriction in the ports themselves it would be about the same loss.

heres the most recent drawing I drew up. It has the reverse return supply and return like Bob Rohr suggested but I drew the circ on the return side to take advantage of cooler temps for the circ. It does show strainers as well but I think I'll just pull the screens after alittle run time.

Here's the boiler side

And the storage side

 

[maple1]

Maybe it's been mentioned already in this thread, but I wonder how much your thermomix is actually doing or needed? It's not like you will have cold storage bottom or radiant floor water wanting to get directly into your boiler. I guess the dT that you will be seeing across your new much bigger HX with this new setup has yet to be seen (unless someone can take a half-educated stab at trying to figure that out). Does your boiler spend much time now sending say <170° water to the HX? How do you control pump launch and what is that set at?

 

[warno]

@maple1

Right now the thermomix does absolutely nothing during normal operation. But if during warmer days when i don't fire everyday the boiler can dip into the 140s-130s and the themomix does it's thing at that point. The pump runs as long as the flue temp doesn't drop below 200° and it's started when I push the "go" button to start the firing cycle. The "go" button turns on the boiler pump, storage pump, and combustion fan. It's all held in the "on" position via a contactor controlled by the flue temp.

And like you said the delta T of the new HX is yet to be seen so it sort of has me worried. On the possibly the delta T isn't crazy low I could just take out the Themomix valve and put in some tees with a manual bypass for season start up process. I'm just worried the deltaT will be lower then I'd like to see so I'd like to keep the throttling affect of the themomix valve.


EDIT:

I Did just look at the price of the danfoss protection valve. It's cheaper and has a CV of 16.2 for the 1.25" valve

 

[BoiledOver]

Since your boiler loop goes through the HX only, your boiler return will be decided by the deltaT. Say the HX delta averages about 20F, you can control the boiler return by controlling your boiler and storage circs with one aquastat set at 165F ON and 160F OFF. This will result in boiler return staying above 140F. The Johnson 419 is affordable and works nicely however you choose to set them up, break on rise or fall, cool or heat. They provide adjustable hysteresis and such.

Notice the link is not to a 419, it has been discontinued and the 421 is the replacement.

 

[Bob Rohr]

I did find a hiccup with the return protection valve. It's a Caleffi Themomix valve with 1" MPT ports. The bigger option for the 1.25" valve is another $200 i would have to spend. So I think I'm going to run my 1" valve for now and if I see issues with it cutting flow to much I'll address it with a larger option. I'll have all the plumbing be 1.25" everywhere else but the valve. I do know this will be the bottle neck, but the 1.25" themomix valve has the same CV of 10 as my 1" valve. so other then the restriction in the ports themselves it would be about the same loss.

heres the most recent drawing I drew up. It has the reverse return supply and return like Bob Rohr suggested but I drew the circ on the return side to take advantage of cooler temps for the circ. It does show strainers as well but I think I'll just pull the screens after alittle run time.

Here's the boiler side

View attachment 221716

And the storage side

View attachment 221717

It is alway wise to avoid any flow restricting devices on the suction side of a circulator. In your case especially, since you are on a fine line of cavitation with pressure/ temperature. Circs are built to handle 225°.

Plenty of other ways to prevent return condensation on the secondary side, as others have suggested. Viessmann and others use a simple 140° control on the return or boiler itself to turn the circ off, a primitive/ bang-bang approach but it doesn't cost you any piping penalty.

In your case I feel you have plenty of HP and return temp protection should not be a big concern. You have bigger fish to fry. It seems you are converting more energy that you can dissipate.

Only on cold start up would you have some condensing. All boilers start cold at some point, the key is to minimize that cold run time. The rule of thumb is to have return temperature above dew point within 10 minutes or so of start up.

 

[warno]

Since your boiler loop goes through the HX only, your boiler return will be decided by the deltaT. Say the HX delta averages about 20F, you can control the boiler return by controlling your boiler and storage circs with one aquastat set at 165F ON and 160F OFF. This will result in boiler return staying above 140F. The Johnson 419 is affordable and works nicely however you choose to set them up, break on rise or fall, cool or heat. They provide adjustable hysteresis and such.

Notice the link is not to a 419, it has been discontinued and the 421 is the replacement.

With that idea would I tie the sensor to the return pipe or water jacket? I'm not sure if turning the pump on and off with the return drop is what I would want or I'm not thinking about it right.

I'm thinking with the sensor tied to the return pipe of the pump shuts off at say 150 there would be no hot water flow to turn it back on. Maybe I'm thinking about it wrong.

My concern with return temps is with my current baby 20 plate HX and low storage temps I'm easily seeing 30° delta T. If I'm going with that 90 plate and bigger dimensions then I would be seeing alot higher delta Ts. I mean with little to no head on the boiler loop i know i will be running alot more gpm which helps delta but the protection valve just seems smart.