avoiding circulator cavitation on unpresurized storage

[pybyr]

for background, I'm going to be using 1200+/- gallons (6x6 ft square, 5 ft high) of unpressurized storage with a flat plate HX

I bought a bronze circulator (Wilo Star 21 BFX 3 speed) (Kind of Wilo's answer to the Grundfos 15-58) intending to use it on the "open" side to pump from the tank to the HX

and then I was reading some of the other threads here about people having circulator troubles, and a mention in particular of how unpressurized OWBs can tend to "go through" circulators if things are set up without an eye to avoiding cavitation, I got to wondering-- 'cause I really want to "do this once, and right" and minimize growing pains.

So I asked my vendor what the NPSH of the Wilo was, and he asked Wilo's tech folks, who responded with this:

"Regarding NPSHr curves for Star, no manufacturer publishes NPSH required for
wet rotor pumps, the reason being we are not concerned as much about
cavitation (boiling water at the eye of the impeller) but more concerned
about boiling the water (lubrication fluid from the system at the sleeve
bearings) in the rotor/can area.

70% of the motor heat is dissipated by the system fluid hence higher system
temps and higher motor heat amperage (larger wet rotor pumps) the faster the
lubrication fluid boils at lower temperatures.

This minimum inlet temp/pressure combination is always a higher pressure
than NPSHr so if the installation meets the "Minimum Inlet Pressure" as
indicated in our technical documents for sure it meets the NPSHr.

For example the Star Minimum Inlet Pressure is:
122 Deg F Min Press 0.7 PSIG
203 Deg F Min Press 4.4 PSIG
230 Deg F Min Press 14.7 PSIG
"

I'm aiming to push the tank to or as close to 200F as I can, for max BTU storage.

but, with a tank 5 ft high, and the circ mounted at the bottom of the tank, if I am doing my math right, I'll have only abiout 2psi of "head" from the weight of the water

so should I be looking at other circs that are non-wet-rotor, and/ or are designed to avoid cavitation even with very low inlet pressures?

if so, any suggestions as to a quality/ reliable/ long-lived one in bronze that won't break the bank any more than needed (I can wish, right?)

thanks, as always, for all the good knowledge, experience, and help 'round here

 

[pybyr]

c'mon gang, please respond.

wood not oil, and the rest of y'all's encouragement to him inspired me to dive in with the flat plate HX, which seems to have tremendous potential-

but this cavitation issue with an unpressurized storage and plate HX seems to have all the makings of turning it into "a beautiful theory, killed by a nasty, ugly, little fact" and I don't want to end up there in the middle of February in VT with a circulator whose impeller blew out from cavitation while I was away at work for an especially long day.......

so, let's solve this detail, I hope

 

[stee6043]

I see where you got the 2psi from. But I'm not sure that is proper math. I think you're missing atmospheric pressure in there. Also, head is measured as a length, not psi (pressure). Calculating head and pressure are two different things.

 

[mtfallsmikey]

Need more specifics...are you mounting the circ. above the water line/fill level or below?

 

[WoodNotOil]

Trevor

Cavitation has me a little worried too. It is not something that came up before I finished my install. My pumps are setup to pull water through my plate hx and I was advised in another thread to switch that so that the water is being pushed through in order to avoid cavitation. I think the tank itself has nearly no pressure by comparison to the drop of going through the plate hx. I will switch mine in the future and you can avoid this by pointing your pumps at the plate instead of away. Also, I went with $70 cast iron pumps on the open side knowing I will eventually have to replace them instead of bronze because of the high up front cost. This is also a good precaution for the unforeseen in this experiment. If I kill a pump in the learning process, at least it isn't a $200-300 one. All I can tell you for sure is that mine works for now. Got the tank up to 171* with an overnight burn!

 

[deerefanatic]

I have the same. I have two pumps, mounted at the bottom fitting facing each other, then the flat plate above them. So, when charging, the water is drawn from the bottom and pushed through the plate, up, and around. But, when discharging, the water is pulled from the top, down the pipe, through the flat plate until I hit the pump, then pushed into the bottom of the tank. I figure if any pump dies, it will be the discharge pump.....

 

[stee6043]

Taking gravity and air pressure into account I figure you should be seeing somwhere around 20psi at the bottom of your tank....

 

[jebatty]

May be wrong, but I don't think static air pressure is relevant, as it is about the same all around, much like a pressurized system - constant pressure throughout the system. I used a circ on an OWB with about 2 psi for 10 years without an issue. It maybe more of an issue if you intend to operate at the extreme end.

 

[machinistbcb]

Another way to help prevent cavitation in a pump is to keep any elbows or tees 12x dia. of the pipe away from the pump.

 

[Nofossil]

I'm going to guess that there's a bit of wiggle room. If they're concerned about water as a colling fluid for the pump, it would be more of an issue at any given pressure-temperature combination if the flow rate were also low. Your flow rate should be very high (low head loss in that loop) so it should see almost zero rise due to heat generated in the pump.

Does anyone know if a cavitating pump makes a specific and identifiable noise?

 

[BrownianHeatingTech]

so should I be looking at other circs that are non-wet-rotor, and/ or are designed to avoid cavitation even with very low inlet pressures?

if so, any suggestions as to a quality/ reliable/ long-lived one in bronze that won't break the bank any more than needed (I can wish, right?)

thanks, as always, for all the good knowledge, experience, and help 'round here

What speed are you planning to run the pump at?

It generates significantly more RPM at speed three than at speed one (1300 versus 2700), so that will have a direct impact on the heat generation, and need for dissipation.

In general, a larger, slower pump will have fewer problems in this realm, so that would be the only recommendation.

Using a drop tube on the outside of the tank to go down to floor level before mounting the pump will help get you as much inlet pressure as you can.

If you were really worried, I suppose you could excavate, install a gasketed box (like is used for sewerage ejector systems), and put the pumps at the bottom, so they would be ~10 feet below the tank water level. I think that might be excessive, though.

I see where you got the 2psi from. But I'm not sure that is proper math. I think you're missing atmospheric pressure in there.

Atmospheric pressure isn't relevant, because it's acting on all sides of the system, not just in one direction. You add it on one side, and subtract it on the other, so it has no net effect.

Also, head is measured as a length, not psi (pressure). Calculating head and pressure are two different things.

"Feet of head" refers to the pressure that is generated at that depth. If you go 1 foot below the surface of a body of water, you will feel a pressure of "1 foot of water column."

One foot of water column equals ~0.433 psi, so Trevor's math is correct.

I have the same. I have two pumps, mounted at the bottom fitting facing each other, then the flat plate above them. So, when charging, the water is drawn from the bottom and pushed through the plate, up, and around. But, when discharging, the water is pulled from the top, down the pipe, through the flat plate until I hit the pump, then pushed into the bottom of the tank. I figure if any pump dies, it will be the discharge pump.....

What's the head loss through the flat plate and piping? It's likely that the restriction of the flat plate will create a lower inlet pressure than would result from having the pump near the surface of the water tank. If the restriction is high enough, you could have negative inlet pressure, rather than the close-to-zero inlet pressure that simply having the circulator near the surface would cause.

Joe

 

[steam man]

Take it from experience-there is a pressure/temperature relationship which can cause flashing in the pump. You will hit a wall at some point. Most likely you are going to have to stay somewhat cooler than you think. You can test Wilo's pump specs from first hand experience and find out what works or doesn't. Slower/cooler helps, hotter and faster hurt performance.

Mike

 

[stee6043]

Hmm...well I'm not one to argue but atmospheric pressure doubles every 33 feet below water you go for a reason. I'm no engineer but I'd suggest 5 feet of water is exerting more than 2psi on the bottom. Although at this point it seems this isn't germane to the conversation anymore....

 

[BrownianHeatingTech]

Hmm...well I'm not one to argue but atmospheric pressure doubles every 33 feet below water you go for a reason. I'm no engineer but I'd suggest 5 feet of water is exerting more than 2psi on the bottom. Although at this point it seems this isn't germane to the conversation anymore....

It's actually ever 34 feet...

34 feet, multiplied by 0.433 psi/ft = 14.7psi

Atmospheric pressure at sea level is 14.7psi

Hence, 34 feet of water column equals 14.7psi, which equals one atmosphere.

The same 0.433 ratio applies to 5 feet, just as it does to 34 feet... and 5 feet, multiplied by 0.433 = 2.17psi

Joe

 

[stee6043]

The problem with your math, Joe, is that air pressure equals 1 atmosphere at sea level. Not at 33 (or 34) feet deep. Assuming you have 33 feet of water with the surface at or near sea level your pressure at the bottom will not be 2psi. It will not be 14.7psi either.

 

[BrownianHeatingTech]

The problem with your math, Joe, is that air pressure equals 1 atmosphere at sea level. Not at 33 (or 34) feet deep. Assuming you have 33 feet of water with the surface at or near sea level your pressure at the bottom will not be 2psi. It will not be 14.7psi either.

You're confusing psig (gauge) with psia (absolute).

If a heating system was sitting in the vacuum of space, then psia would be the correct unit of measure. We don't tend to install wood boilers in space, so we measure pressure relative to atmospheric. In other words, 0psig = 14.7psia.

I realize this can be confusing, but you just need to remember that gauge readings are relative to the atmosphere, not an absolute vacuum.

If you go down 34 feet from sea level, you will be at 14.7psig, or close enough to it (temperature changes water density, so small variations are possible). If you go down 5 feet, like in the case of Trevor's tank, you will be at just a bit over 2psig.

Joe

 

[deerefanatic]

That is correct..... Joe IS right here.....

 

[pybyr]

Need more specifics...are you mounting the circ. above the water line/fill level or below?

my tank has 2 ports about 6 inches above the bottom.

I will tie both the inlets and outlets from the tank to the pump "through" those, and with piping inside the tank- to my large-diameter "stratification baffle" which will both pull and push flow using the pump from the bottom of the tank.

So the "head" of the height of the tank will be above the pump in both flow directions, with low velocity flows within the tank

so I should get the approx 2 psi from the weight of the water in both flow directions

I don't happen to have the specs of the 5x12 inch 70 plate FlatPlate HX with me (on the road for work) in terms of back pressure through the HX, if that may affect things, but one of the reasons that I went with the 70 plate is to minimize back pressure/ maximize heat transfer

anyone know of any bronze circulators that are likely to be really content with this set of parameters of lots of flow and really little NPSH? Maybe a non-wet-rotor-type?

 

[pybyr]

[quote author="BrownianHeatingTech" date="1226022871

If you were really worried, I suppose you could excavate, install a gasketed box (like is used for sewerage ejector systems), and put the pumps at the bottom, so they would be ~10 feet below the tank water level. I think that might be excessive, though.
Joe[/quote]

I'm a non-engineer who wishes I was an engineer, and so tends to follow the "overkill is just enough" approach, so that's an interesting idea

but if I blew out the concrete cellar floor and went more than a foot or two down, I'd quickly be into the groundwater table...

really... the old timers who built my house around 1830 found the one spot in the 100+/- acres that was then the farm (of which I still own 15 that prior owners hadn't carved off) found the one and only spot in the whole property that could actually accomodate a basement without running into shale ledge, granite outcrops, or perched water table (I still wonder exactly how they figured the spot out)- but there is _not_ much room to go deeper without getting down into the groundwater

 

[deerefanatic]

Them old-timers were smart..... Just like witching for water.... It's proven that it works... And not one person can tell you how or why.......

 

[WoodNotOil]

I think the whole discussion may be getting way too technical and your question seemed more like a practical one. You want to avoid cavitation. As far as head pressure is concerned, your flat plate is the largest contributer. Mine was calculated to have 12 feet of head. Add a few twists and turns to that and you have rough estimate. The feet of head probably has more to do with sizing your pump than with cavitation so long as you follow the guidelines of pointing your pumps at the flat plate and making sure there is 12x the diameter of pipe distance to anything on either side of the pumps and that should give you the practical advice you are looking for. I assume a pump intended for hydronic use can withstand the heat of pumping really hot water through it, otherwise the company would go broke replacing them under warranty. The weight of the water will help with priming your pumps and probably has little effect on feet of head pressure. Is that practical enough for you?

 

[Couderay80]

The best way to get away from cavitation is to look up a piece written by Noel Murdough Zone Heating with Condensate Google Him

 

[pybyr]

Them old-timers were smart..... Just like witching for water.... It's proven that it works... And not one person can tell you how or why.......

my grandfather, who made his living keeping the accounts-receivable for a large metals company, even though he lacked a degree in the field, could witch for water, and people sought him out to have him do it

when I was a kid, way before I could even understand what it was about, my dad gave me a fresh-cut forked cherry stick and let me walk around the yard, and sure enough, when I walked over the septic or the (buried) well, the stick practically pulled out of my hands. it also went wild down in the parts of the back yard about 20 feet uphill from some "seeps" from springs

I tried it again in my 20s after getting an "education" and came up a total dud.

still trying to figure that one out, or how to re-gain the knack, and in the mean time, it seems like a sad reflection on something

 

[pybyr]

The best way to get away from cavitation is to look up a piece written by Noel Murdough Zone Heating with Condensate Google Him

thanks- I pulled the article up and it looks like a great resource, but I am too fatigued to absorb it all tonight- but thanks again-

 

[pybyr]

I think the whole discussion may be getting way too technical and your question seemed more like a practical one. You want to avoid cavitation. As far as head pressure is concerned, your flat plate is the largest contributer. Mine was calculated to have 12 feet of head. Add a few twists and turns to that and you have rough estimate. The feet of head probably has more to do with sizing your pump than with cavitation so long as you follow the guidelines of pointing your pumps at the flat plate and making sure there is 12x the diameter of pipe distance to anything on either side of the pumps and that should give you the practical advice you are looking for. I assume a pump intended for hydronic use can withstand the heat of pumping really hot water through it, otherwise the company would go broke replacing them under warranty. The weight of the water will help with priming your pumps and probably has little effect on feet of head pressure. Is that practical enough for you?

the good tentative news from Wilo, by way of tech support from Patriot-Supply (both of whom have been REALLY patient with my questions on all this) is that apparently there is indeed a significant margin for error in their design assumptions.

Apparently, according to them, I'll be running outside of specified norms, but not in a way that they think I will generate any immediate mushroom clouds, and they're seemingly willing to work with me if something goes bang in a very short period of time.

Wilo seems to have put a lot of thought into their volute and impeller designs (they do, after all, build nothing but pumps, and are the biggest pump company in the world) so I guess, for the moment, I'm leaning towards rolling the dice and seeing what happens.

anyone have any suggestions on how to detect cavitation-in-progress before it leads to destructive results? do you listen with a mechanic's stethoscope? could you strap a sensor on the pump and hook it to an oscilloscope and watch the patterns and frequencies? other techniques?

thanks