Hot water recirculation from atmospheric tank

[Oli_eng]

Hello everyone,

It's my first post here so my apologies if I'm in the wrong section.
I am designing the plumbing of an industrial plant and would like your opinion on the following concept:

I have a hot water storage tank at atmospheric pressure with booster pumps to pump it to the plant process.
There is a fair amount of distance between the tank and the point of use and the process doesn't use hot water all day long so there is periods of time where the hot water just stand still in the pipes. So in order to have hot water ready when the valve open I have to recirculate somehow.

Should I just plug a smaller pipe at the end of the main line and connect it to the atmospheric tank and let the booster pump run 24/7 or is there a better solution?

Thanks for your help

 

[kjoiner]

I far from an expert on this subject but would a recirculating line work. I'm thinking something similar to what people use in home to ensure a quicker supply of hot water to a faucet that is far from the heater.

Kyle

 

[LittleInch]

You have two options.

I assume first off that the hot water line is insulated?
How hot is "hot"?
how much variance in terms of temp will the plant accept. So e.g. if your tank is 65C, will 60C be ok? Or 55?

You can as you say and have a return line - often this will include a small pump at the end point to pump the fluid back to avoid running the larger pump all the time. I think the ROT for these is about 5% of the normal flow needs to be recirculated to keep it within a few degrees of whatever the required temperature is.

Or so long as the insulation is effective, add am electric heat tracing cable, either self regulating or simple on/ off with a temp control.

Or a different version using steam or any source of waste heat your plant might have as a trace heating fluid.

PS - would probably have been better in piping or heat transfer forums, but this one will do for now.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[EdStainless]

Flow in the 5-10% range is enough to do this.
So if you had a 4" supply line you could use 1" or maybe 3/4" as a return line.
Make sure that all of the line is insulated, even the small return part.
I have seen people just go to the end of the header and connect a small return pump there.
Watch your return line. Continuous flow of hot water can greatly accelerate erosion and corrosion issues.
Since this is an open tank, unless it is treated with buffers the pH will be acidic from CO2 being absorbed.
Granted in hot water the amount of gas absorbed is much less, but it still happens.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Oli_eng]

Thank you all for your answer.
Everything is insulated and maybe it will be enough to keep the water warm but I don't want to take any chance, especially if for any reasons the water stand still for too long and end up loosing all the heat.
I thought about a recirculation pump but is it useful if my tank is atmospheric?
Let's say that the booster pumps pressurize the hot water line at 90psi all the time, if I plug a 1" recirculation pipe from the main line to the atmospheric tank, wouldn't the water flow 'naturally'? The booster pumps are VFD, so it would just run on low speed to keep the pressure at 90.

 

[LittleInch]

A recirc pump is better for that sort of flow as even a good VFD will struggle to be efficient at a 20:1 turndown, but depends on how many pumps you have or need. Far better to just install a little recirc pump either next to the main pumps or right at the end.

Which option works best depends a lot on how often the line runs to the main user. Anything less than 20 to 30%, trace heating might be a better solution or at least considered. Power supplies and electricity cost ight make it not preferable, but then a recirc lines needs a pump and duplicate length of pipe, even if its a lot smaller than the main line.

I don't know what you mean by "naturally"? Can you explain a bit more as it won't circulate by thermal gradients and will need to be pumped.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[shvet]

this may help

https://www.eng-tips.com/viewthread.cfm?qid=501898

 

[Oli_eng]

Here is a badly drawn picture to better explain myself.

Let's say the ball valve at the process is closed and the main pump is always pushing 90 psi.
The fact that the main line is connected to the atmospheric tank (at 14.7 psi) through the recirculation line would mean that the water just flow around because of the pressure difference right? So I would assume that a recirculation pump is not useful.

Now I agree that using the booster pump to just recirculate the water is overkill and not energy-efficient and as you said LittleInch it would not even work.
So I would need some kind of controls to start the booster pump when the pressure really drops a lot (i.e the process valve is open) and stop it when the pressure goes back to a certain threshold. Then I would start the recirculation pump to always have hot water in the main line.

The tricky part is that the pressure sensor for the booster pump has to be adjusted to take into account that the main line is always connected to the atmospheric tank, and so the pressure is never 90psi. Should I put a balancing valve on the recirc line to minimize the pressure drop?
I don't know if it makes more sense now, but I appreciate your feedbacks.

Thanks

 

[EdStainless]

Stop, get out your pump curves (shoeing various frequency of operation) and your VFD info.
If you run this pump at 1/20 of full flow will it function? Will it generate enough head to flow any return back through the 1" line?
Alternately, if you turn it down until it will just barely flow in the return line what will the flow rate be?
If you don't get nice answers for these questions then you can't just use the main pump and will need a small recerc pump.
The recerc pump flows are so small that these pumps run all of the time, even when there is demand.
The have such minimal impact on the system.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[LittleInch]

OK, badly drawn it might be but it is not correct.

Thinking about htis some more, your recirc pump really needs to be in parallel with and with NRVs to your main "booster" pumps.

where you have put it or even at the connection with the main line could easily lead to lack of NPSH for your pump as there is only atmospheric pressure plus what ever head difference there is between liquid level in the tank and the pump inlet flange.

your pressure at the bottom of the tank IS NOT 14.7psi, unless the level in the tank is always 10m above the exit?

I think you mean either 0.2 psig or 14.7 psia?

You seem to be confused about the atmospheric pressure thing and where you need to add differential head from the pump?

A pressure switch by the valve or valve position could both be used to signify a start procedure, but there will always then be a time lag and a bit of a surge as the big pumps kick into life. You might need to add an accumulator or pressure tank somewhere in the system to smooth out the start / stop pulses.

Also you will probably need to add a control valve on your return line to stop too much flow coming back down the recirc line.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[GregLocock]

Sorry Ed, complete brainfart.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Oli_eng]

I'm going with the recirc pump solution as suggested by LittleInch, thank you all for your replies, really appreciate it!