Underground pump vault airflow considerations

[Jos123]

Hello,

I am currently designing and building a small utility vault to house a centrifugal pump for a water fountain, and I am specifically concerned with providing adequate ventilation not only during operation but also in general to reduce moisture and corrosion of the equipment. While all of the equipment is on a small scale and so the stakes are not high, I am hoping to avoid certain issues.

The vault is to be 22 inches deep, 22 inches wide, and 24 inches long. The walls will be constructed of concrete, approximately 8 inches wide. The vault will be covered with a concrete lid, fitted into a recess in the concrete wall, and while not water/tight, will be tightly fitted enough.

The pump has a 1/2 hp. TEFC motor. The pump is to be mounted on its own concrete pedestal, roughly 5 inches in height, providing a clearance for any possible water intrusion and to reduce pump vibration, etc. At the rear of the pump, below the pedestal and through the wall is a 3 inch pvc pipe opening which serves as both an overall scupper style drain and an incoming air supply. This 3 inch pipe will either tee off from its main drain line and surface right outside the vault, flush with the ground with a grate, or it will simply rely on air via the drain piping itself, as this quickly increases to a 6 inch pipe. The second option may allow me to pre-condition the air somewhat and so provide better air exchange in the vault. Its effectiveness is debatable, though any sort of pre-condition may increase performance. At the moment, however, this is not necessarily my main concern.

My main concern is with the exhausting of hot air from the vault and general air exchange. I do not think such a setup warrants any extractor fan. Besides, I want a passive design for when the pump is not in use. The simplest solution would normally be at least an equal size (3 inch) pipe positioned in front of the pump as high up on the front wall of the vault as possible to promote a heat exchange. However, I am concerned about having such a large pipe (which will also be foam wrapped) in the concrete, with only several inches of concrete mass above it, as I feel this might lead to cracking. If I shift the exhaust vent pipe lower, I am worried I may lose the heat exchange effect.

One possible solution is to split the exhaust into two smaller pipes, spaced evenly in front of the pump. The smaller the penetrations, the safer I feel it would be to position the exhaust higher up and closer to the concrete edge. In additional, two pipes spaced apart may also better match the direction of the hot air off the pump, as the air will blow over the motor cooling fins, hit the pump housing itself, and likely disperse sideways somewhat, rather than straight out in front of the pump (the only exception to this being the hot air from under the pump motor, as the pump housing will not block it here). In this way, the hot air may find its exit more quickly.

The question is how does the use of two exhaust pipes actually perform. The first concern would be to match the cross section size of a 3 inch pipe. Two 2 inch pipes, for instance, would still not add up to a 3 inch in cross section, but perhaps be adequate enough, considering the possible benefit of a higher positioning in the vault. Assuming I match the cross section sufficiently, the concern then would be the increased surface area of the double pipes and so the added friction. And if I understand correctly, having multiple pipes means the air will have a larger exit area and therefore the air flow will slow down.

The objective of course is to move as much of this hot air out as quickly as possible. So, will two pipes no matter what still fail to generate the same suction effect of a single pipe, even if that single pipe will have to be positioned lower? Or is there a way to manipulate the exits of the two pipes to compensate for the slower airflow, for example by reducing the pipe diameters before exiting to open air?

I should note that my intention is to make the exhaust flush with the ground, immediately outside the vault, and covered with grating. (Any attempt to have this pipe stand above the ground would involve running it at least 15 or so ft. away, in order to conceal it, which would be worse I'm sure than exiting at ground level.) To allow for any water accumulation, I plan to use a sanitary tee fitting (or perhaps even a regular tee), similar to a plumbing vent, and this will be tied into the drainage system. So, the exhaust pipe will not be a direct flow to open air exactly, but I'm not sure I have any other option. How this affects the air flow, I'm not sure. Perhaps it would just react similar to an an elbow.

One additional note is the sizing of the 3 inch incoming supply pipe is somewhat arbitrary. I basically chose to match approximately the motor's fan size, which is under 4 inches. A 4 inch pipe seemed excessive.

Anyway, I hope I've given a clear enough picture of the project. I understand a lot of this is theoretical and probably would have to be tested actually to determine the best approach. But if anyone has any information or advice, I'd appreciate it very much.

Thanks,

Joe

 

[256bits]

Welcome @Jos123

By what method are you choosing the motor enclosure?
As a guide,

​

 

[Jos123]

Hi 256bits,

Thanks for the reply, and Happy New Year!

The enclosure's main function is visual concealment, as well as protection against general access to the equipment/electrical and general weather conditions.

The fountain will be winterized, and the pump most likely will be removed and stored, so there'll be no concern for operation during iced over conditions. As the motor is basically a sealed unit, ingress of dust, etc. is also not an issue. Prolonged or even temporary flooding, though, would be a problem for the motor, which is one of the reasons for a sufficient pump pedestal height and for the 3 inch drain/wet vent. A further safeguard to flooding will be the extensive use of gravel for its subbase and backfilling.

Thanks,

Joe

 

[Nik_2213]

Have you considered a balanced ('Coaxial') flue arrangement ? Such that taller, inner channel vents upper part of vault, while outer feeds lower ? Hot air rises from top, replaced by cooler, denser air at base...

Hide stack by integrating with robust ornamental feature such as trellis, arch or arbour. 'Down-spout' mini-planters may clamp onto such pipes, suit hanging flowers etc..

If you have cat(s), craft a 'jungle gym'...

FWIW, given cost of rest of system, a small, float-switched sump-pump would seem affordable protection. If discharges above ground, a vault soak-away is not required, off-setting sump-pump costs...

 

[Jos123]

Hi Nik_2213,

Thanks for the reply. I actually did not consider a coaxial flue. It's an interesting idea to minimize wall penetrations. Would this be less effective than an air supply directly at the pump's fan, though? Or perhaps that is just a simple way of looking at it, and an air supply is an air supply.

Regardless, I probably still can't make use of a coaxial flue because the area around the water fountain is going to be quite minimalistic, with only pea gravel and some low hedging. It's a reflecting pool style fountain. If I were to try to conceal a stack, first, it would have to be a shorter stack, in the range of 1.5-2 ft max and then it would have to be about 10-15 ft away. This additional distance, with the additional fittings involved, would, I think, outweigh any benefits of the added height, especially with not much added height.

As for the small sump pump, yes, it's a possibility, depending on how inexpensively that could be set up, relative to the cost of the main pump itself. My hope, though, is that the 3 inch drain would be sufficient. This 3 inch drain would within a short distance, between 5 and 10 ft away (I haven't yet worked out the overflow and other drain tie ins), increase to a 6 inch pipe going to daylight.

Thanks for the input.

Joe

 

[Jos123]

Hi everyone,

I am thinking that I should probably further analyze my airflow/air exchange requirements for the underground vault, and perhaps I am getting ahead of myself in trying to manipulate the piping location and design first. In this regard, there may be some more concrete information I can follow. I was uncertain whether to start a new thread, but I suppose for now I'll include it here.

So to begin with, the pump is a 1/2 HP, .37 kW, and it is rated at 104 degrees F. The outside temperature should not exceed about 95 degrees. If I do precondition the air, at least somewhat via the combined drain/vent (acting like an earth tube), then I expect to be able to bring this temperature down to something around 80 degrees F. I think that's a fairly conservative temperature to achieve.

If I did the math correctly, then the motor itself should require roughly 50 CFM with pre-conditioned air and roughly 130 CFM for un-conditioned air.

However, how does the motor's actual fan factor in to these numbers? And, with the motor's fan operating in a confined space (22" wide, 24" long, and 22" deep), is the air supply still considered passive?

I know for passive duct working there are CFM numbers to follow, but I don't know if this applies exactly. Also, I understand for mechanical rooms, etc. there are usually some guidelines for vent sizing, such as 1 square inch/1000 BTU. Again, I do not know if some such guideline applies to my scenario.

Anyway, if anyone has some thoughts on this, I'd appreciate it. Also, if you think this should be explored in a separate thread, please let me know.

Thanks.

 

[Tom.G]

For a 9 degree F rise at 420W (370 for pump + 50 for fan), the calculator below shows that 140 CFM is needed. That is roughly a medium size bathroom exhaust fan that uses a 4 inch vent size. Of course that does not include any safety factor, I would suggest at least a 50% increase to 200+CFM.

Also be aware that this estimate also assumes that ALL the circulating air flows directly to the motor!

You may not want to operate at the motor temperature limit -- increase the fan CFM as needed for a lower temperature rise.

If you have a long run or several bends, you will need a higher fan CFM rating to overcome the flow resistance -- or a larger diameter vent piping.

https://farnam-custom.com/resources/calculators

Cheers,
Tom

 

[Jos123]

Hello Tom,

Thanks for the reply and the link to the calculator.

Regarding the motor's fan, sorry if I wasn't clear about this, but the pump and motor is a single fan cooled unit rated at 370 watts, with the fan at the rear providing the only cooling and so otherwise sealed. When you say an additional 50 watts for the fan, I assume you meant if I had a separate fan installed in the vault to provide air.

My hope was to avoid any additional extractor or circulating fan, since my piping will be as short as possible on the exhaust, within 1 foot. The distance would be similar for the intake if I do not precondition or farther if I do, with the trade off of the air entering cooler. The intake also is to enter the vault at floor level, which would be only 4 inches back and only about 4-5 inches below the pump's fan itself (pump is on a small pedestal).

Anyway, if you could let me know how the motor's own fan factors in, I'd appreciate it.

Thanks again.

Joe

 

[Tom.G]

Anyway, if you could let me know how the motor's own fan factors in, I'd appreciate it.

The motor built-in fan is to circulate the ambient air thru the motor for cooling.
The assumption is that the ambient air is cooler than the motor's temperature limit.

Unless you duct that fan to outside the enclosure, you will have only convection to keep the enclosure cool. (along with some thermal conductance thru the enclosure walls)

Besides, I want a passive design for when the pump is not in use.

You could wire a fan to the motor so it comes on only when the pump runs.

Please realize that the minimum 200 CPM air change rate with little safety factor means that the air in your 8 cu.ft. enclosure has to be completely removed and replaced 25 times per minute (about every 2.4 seconds). This will NOT happen with natural convection with a few feet of 3 or 4 inch pipe.

If you can still get some 60W to 100W incandescent light bulbs, put 370W of them in a cardboard box the size of your enclosure, close the box, cut a couple 3 or 4 inch holes in it, power them on., and stick a thermometer in the box.

Oh, another thing I just realized:
It is unclear to me if the enclosure lid is at ground level. If it is, and in direct Sun, there wil be an additional 30 to 70W from the sunlight shining on it.

It sounds like you will have a pleasant backyard environment upon completion.

Cheers,
Tom

Random thought:
Can you relocate the pump to above-ground, preferably shaded, perhaps concealed with some plants? That would certainly eliminate the ventilation problem.

 

[Jos123]

Hi Tom,

Thanks for your continued input, I appreciate it very much.

I guess my hope in positioning the 3 inch intake pipe in vicinity of the pump's fan was that, although not ducted to the outside directly, the fan would have immediate access to cooler air, especially if pre-conditioned. I considered positioning this intake exactly at the location of the fan, nearly ducting it directly as you've mentioned, however, I wanted the vault to vent naturally when not in operation and so chose to lower the intake to maximize air exchange. Also, I am aimed to minimize concrete penetrations by having the intake double as a drain when necessary.

So, if I could redesign the intake to ensure cooler air to the pump fan, by means of an earth tube style design where the piping runs deeper in the ground and over a sufficient distance, perhaps this would help to drop my CFM requirements substantially. Rather than max 95 F air entering the box, maybe I can drop the temperature to the 70s F.

Now, I'm not necessarily opposed to adding a separate fan for the vault, and wiring it accordingly, I just again hoped to manipulate the air intake and particularly the exhaust in some way to rely on convection. But without a substantial stack for the exhaust, as you say, such suction might not be possible.

The max size pipe penetration I am comfortable with for the exhaust is a 3 inch pipe, based on its vicinity to other necessary pipe penetrations and the concrete edge itself. A small 3-4 inch axial fan would only exhaust about 40-50 cfm, though. This could work in combination with a reduced intake air temperature of 70-80 F. Otherwise, I'd have to consider sizing up the fan and reducing the ducting through the 3 inch penetration. That's not ideal, to reduce it, but I suppose possible.

Another possible option is to vent the exhaust through the lid itself. Such a setup would allow water ingress from rain, etc. Now, I could somewhat direct this water so as not to drip over the more sensitive components of the system. I could also try to make use of a low profile mushroom vent as used on boat decks. In the end, the increased ventilation may dry out the vault quickly enough so as not to be much different than a lesser vented vault that is still only moderately sealed off from water intrusion.

It's not really my first choice, but a vented lid would allow the greatest degree of convection, as it would of course be the highest. I also have more flexibility in sizing it up, as the lid will be quite reinforced to begin with and is far more easily replaced should it crack over time.

This would be the closest I could achieve to an open air pump setup. Unfortunately, I am already committed to its location close to the basin for flow reasons and because I do not have the space in the area to conceal it above ground otherwise.

And yes, the vault cover will basically be at ground level and in the sun. I may superficially cover it with pea gravel, and that might help somewhat with the radiation. If not covered, it may get whitewashed to blend better, but yes there will be some additional heat to consider.

I think I am going to take the time to test various setups as you suggest with lightbulbs, etc. That's a good idea. My hope (or pipe dream...) is still to figure out some method to condition the incoming air enough and generate enough exhausting without additional energy consumption. If there is anything else unconventional that might work, I'm all ears.

Thanks again.

Joe

 

[256bits]

Regarding the motor's fan, sorry if I wasn't clear about this, but the pump and motor is a single fan cooled unit rated at 370 watts, with the fan at the rear providing the only cooling and so otherwise sealed. When you say an additional 50 watts for the fan, I assume you meant if I had a separate fan installed in the vault to provide air.

Does the motor give off 370 w of heat, or does it provide the capability of up to 370 w of power to the to the pump,
If it is giving off 370 w of heat you have an excessively inefficient motor, or running excessive load near stall.
Use a more a reasonable value for heat generation, perhaps 5% - 10% of rated capacity.
Or, find out the actual running point of the motor, and use use the efficiency at that point for heat generation.
Or, the power PQ ( pressure, flow rate ) of the water that the motor is generating.

I would for a passive heat sink design, which I think you already have in mind,
Less things to go wrong, and cause havoc and headache.
The water running through the pump is itself a heat sink and will provide some cooling. As well, the walls of the enclosure will conduct some heat away from the enclosure as long as the soil temp is less than enclosure temp, some depending upon soil type surrounding the enclosure.
Your soil type,

A further safeguard to flooding will be the extensive use of gravel for its subbase and backfilling.

You should do a simulated field test:
Pump some water through the fountain nozzle(s) with a mock setup.
After some time, note the motor temperature stabilize, when,
- pump and motor are in open air
- motor is in an enclosed box. <-- no convection
- pump and motor are in an enclosed box. <-- no convection, pump acts as heat sink
- pump and motor in enclosed box with holes for passive ventilation. <-- convection, heat sink

Compare stabilized motor temperature with ambient air temperature.
Do you have cause for concern from the data obtained?

 

[Jos123]

Hi 256bits,

Thanks for the reply and for the heads up. Yes, you're right, I was providing the rated capacity, it seems, for the motor, not the generated heat in wattage. Sorry for the confusion.

So, I checked the pump spec. sheet further. The nominal power is 370 W and the absorbed/input power is 650 W. So, if I'm calculating correctly, that puts the motor efficiency itself at about 57%. Then, using the electric motor heat loss calculation, the motor should be giving off about 544 BTU/hr.

Now, if this information is correct, and assuming I do not precondition the air, and my temperature change is only about 10 degrees (104 F operating temperature from probably max 95F), then I should be looking at a necessary 50 CFM, give or take. Does this seem accurate, and if so, how do these numbers sound for a passive, convection setup?

As for the actual running point of the pump, that's not quite as simple for me to say. At the moment, I do not have all of the details sorted out, such as fountain nozzle style and size (whether it will simply discharge into the basin or if it would have a jet feature positioned just at the water surface, etc.). Furthermore, this pump will be serving two purposes. The first and primary purpose will be for the fountain circulation. The second will be to power a small irrigation zone with a couple heads (the exact number and distance to be determined), allowing me to frequently change the fountain's water, rather than relying on chemicals.

As a result, the pump I selected, a .5 HP with max capacity of 66ft of head and max volume of 26 GPM) should be somewhat overpowered for its main function and probably require some degree of throttling. If I've done the math correctly, the water power required for the fountain operation itself is more in the range of .22 HP, putting the efficiency at about 44%.

I do not know, though, how this pump efficiency point translates to the heat generated. Running on the right side of the curve should allow for a large exchange of water through the pump, which I assume would help with heat transfer.

I had not actually considered the heat sink effect of the pump itself in terms of overall ventilation. Testing that in various scenarios would be interesting to do. Though to match the exact operating conditions, it seems would require being hooked up to the fountain's plumbing beforehand, otherwise the suction head and discharge pressures would vary and with it I assume the heat generated in the space. But, my plan is to build the vault and plumb through it afterward, for ease in building the form work and placing the concrete and a few other reasons.

Anyway, thank you for your input. Hopefully, I've provided more accurate information this time.

Thanks,

Joe

 

[256bits]

As a result, the pump I selected, a .5 HP with max capacity of 66ft of head and max volume of 26 GPM) should be somewhat overpowered for its main function and probably require some degree of throttling. If I've done the math correctly, the water power required for the fountain operation itself is more in the range of .22 HP, putting the efficiency at about 44%

pump efficiency == actual shaft power input to pump / water power output ( ie PQ ) to the piping
Is the the 44% is the efficiency that the pump is operating at from the pump data sheet and curves.

As for the actual running point of the pump, that's not quite as simple for me to say. At the moment, I do not have all of the details sorted out, such as fountain nozzle style and size (whether it will simply discharge into the basin or if it would have a jet feature positioned just at the water surface, etc.). Furthermore, this pump will be serving two purposes. The first and primary purpose will be for the fountain circulation. The second will be to power a small irrigation zone with a couple heads (the exact number and distance to be determined),

One needs the system curve to match a water pump, and then select an electrical motor to drive the pump.

We are getting away from the heat buildup in the container, to a design of the whole system.

I did say this

Use a more a reasonable value for heat generation, perhaps 5% - 10% of rated capacity

erroneously assuming the electric motor would be under light load.

If you say

the motor should be giving off about 544 BTU/hr.

or 160 watts.

Though to match the exact operating conditions, it seems would require being hooked up to the fountain's plumbing beforehand, otherwise the suction head and discharge pressures would vary and with it I assume the heat generated in the space

You are testing the mock enclosure for heat dissipation at various heat generation conditions.
The exercise is to indication of whether forced air flow is necessary to cool the space, or whether passive is adequate for a particular heat generation within the enclosure and temperature rise of the air.

Use light bulb if you wish to represent the motor for the heat generation.
Use a cardboard box to put over the lamp light bulb.
Vary the input and output air hole location.

Otherwise, attach a valve to the pump to vary the pressure and output flow.
Put the box over the pump/motor.
BY varying the valve opening, the electric motor will be under different loads at different PQ, and generate different heat values.
If the mock is tested at various points, somewhere along the way, the pump/motor will be functioning at the exact operating conditions, or near to it by extrapolation.

 

[Jos123]

Hi 256bits,

Thanks for the continued help.

The pump is an over the counter option, so to speak, given the scope of the project. So, I matched curve as best as possible with my requirements and plumbing conditions, knowing that throttling may be necessary for general use.

Yes, the 44 % efficiency operating point is based off of the curve data provided by the manufacturer (Speroni, model cm 22), at lowest pressure and max flow, which is 32.8 ft of head and 26.4 GPM. So, Pump Efficiency= ((26.4 x 32.810) / 3960) / .5 hp.

But, yes, as you've said, we're getting into a larger topic than intended. I was just curious as to how running on the right of a curve would influence the heat sink capacity of the pump itself, as related to the overall ventilation conditions. I assume it would be beneficial.

As to the 544 BTU/hr. of heat generated, yes, that does seem to be a lot. I am basing that number off the rated capacity and Input power listed. So, motor efficiency = .37 kW/.65 kW or 56.9%. Then, Electric Motor Heat Loss = .37x (1-.569) x 3412.142.

You're right, I can test the pump at various discharge pressures and see how it reacts in an enclosure, etc. Or simply use whatever generates equivalent heats and go from there.

My original intention was to find out what ways the intake or exhaust piping could be manipulated to increase convection effects, speed up airflow, etc. But, first, I thought I should find out if I was accurate with the CFM requirements. Anyway, I apologize if I've crammed too much into this thread topic and veered off course.

If you think there is anything additional to consider at this point, other than performing various test conditions, I'd appreciate your ongoing input.

Thanks, everyone.

Joe

 

[Tom.G]

During your testing, the use of a Kill-A-Watt style plug-in power meter could be worth the small (USD $10-30) investment.

Cheers,
Tom

 

[Jos123]

Okay, I'll look into it and see what can be gotten cheaply. Thanks for the tip, Tom.

Joe