pressure tank derate or eliminate

[bentov]

An irrigation supply system for a large cemetary has (3) 30hp split case pumps & a 4,000 gallon steel tank (w/AVC system, lead/lag controller, etc.), formerly used to supply & pressurize a sprinkler/drip system at the end of a 1.2 mile pipeline.

A redesign now has a lake at the end of that line, with a new booster pump station drawing from the lake for system pressure; the booster controls open & close a lake fill valve at the end of the original pipeline. We're going to change the supply pumps to (3) 15hp units for better performance at the lower head condition. Meantime the 30's are running with discharges throttled to prevent cavitation, have adjusted the pressure controller so they're either on (when valve is open) or off (when closed), not really cycling like a normal pressure system anymore.

We have no level communication between stations so the plan is to leave the controls this way, question is what kind of tank capacity do we need now? Can we ditch the big tank & AVC, put in a small bladder tank instead? How do we size that?

 

[BigInch]

You need to find the maximum drawdown in tank level during expected minimum and extreme demands verses maximum and minium supply (respectively). It seems like the lake has replaced some capacity of the tank, but without the exact configuration and expected demand and supply numbers, hard to say what you can eliminate.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[bentov]

Thanks for your response . . .

I was able to measure flows & pressures with the exisiting pumps, selected new pumps based on observed losses plus known current & future system demand (in gallons per day drawn from the lake). One new pump running will do 470gpm @ 89hd-ft(39psi)with the fill valve open, should take care of current demand while running only during off-peak utility rates. As demand grows, depending on power rates, etc., we may run two pumps which will produce a total of 728gpm @ 122hd-ft(53psi). Though unlikely, running all three pumps together would result in 819gpm @140hd-ft(61psi).

In any case, the setpoints will be slightly below the valve-open running condition for ON, slightly above it for OFF (so something like 35psi ON, 42psi OFF in the one pump scenario, no sense wasting energy to build pressure any higher). I'm thinking it's similar to the constant pressure VFD situation, where only a very small bladder tank is needed sort of as a reference . . .

 

[BigInch]

Sounds like you have done the homework on the demands, but you don't say anything about supply. What is the capacity of the supply? Can you supply a constant 470 gpm, 720 gpm and 819?

The way I see the system diagram is,

[supply source] -> [tank] ->
[3 x 15 HP] -> [Pipeline] -> [Lake] ->
[New Booster]-> [irrigation system] -> 470/720/819

Tank_Outflow = pipeline's flowrate (gpm) x pipeline operating time (minutes) = Tank_Outflow (gallons)

Tank_Inflow = supply flowrate gpm x time of availability (minutes) = Tank_Inflow (gallons)

Tank Capacity = Tank_Outflow - Tank_Inflow

You can deduct volumes when pipeline operating time overlaps the supply availability time.

Basically, to size tanks you must integrate supply to that tank with demand from that tank (or a lake). That's just keeping track of the volume in - volume out. Tank size is found from the difference between max and min accumulated volumes.

http://files.engineering.com/getfil...504-41fa-92b7-f43d4e76ff2a&file=TANK_SIZE.png

You can use the results to size the pipeline assuming various operating scenarios, such as this one which assumes a constant pipeline flowrate, with the pipeline running either on or off. If you have a variable pipeline flowrate, just change the pipeline's flowrate to suit your variable rates.

You can examine operating methods to optimize the configuation. For example, it might be more cost effective to have a constant pipeline flowrate 24 hours a day, which could require smaller and fewer pumps, perhaps a larger tank, but the lesser cost for pumps, plus the ease of operation makes up for the cost of the tank, or other operating cost scenarios of your choosing.




"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[bentov]

Thanks again for your attention . . .

The demand side is the 3 million gallon lake, approximately 1-2 foot fill valve operating differential. The supply is the 3 new pumps, running at constant volume only when the fill valve is open - given the pipeline losses, one only running produces the 470gpm (and supplies enough to keep the lake full at current demand levels running off-peak). So, the new supply pumps will be held off by a timer for rate management, then operated by setpoints on the digital pressure controller. Initially we'll use LEAD only and alternate primary duty.

If the lake level starts to fall behind, we'll decide whether to stay in the single pump mode and disable the time of use controller (for longer run time), or go with 2 pumps running by enabling LAG1 at the same setpoint (which results in the 728gpm flow at the higher head due to higher volume in the same pipeline) and stay on time of use - just a money thing there, more watts per gallon at the higher head but still maybe cost effective given the rate structure . . .

In any case, the big tank is just tee-ed off to the side, water doesn't have to pass through it. I think I understand the tank sizing you describe, but isn't our situation different? We don't have the inflow/outflow - we'd rather be using lake level setpoints instead but no ready link so using the existing pressure controls for convenience/cost saving. I'm thinking the tank needs to supply only enough air cushion to prevent the pressure oscillation that occurs when there's no air available, just not sure what that cushion volume needs to be . . .

 

[BigInch]

I doubt you need it then. You're thinking of a pressurized water supply system for a house supply, where you have a bladder tank that stores a certain amount of pressurized water when the faucets are closed, and when opened, water runs from that tank for some time before the pump needs to turn on. The pump is usually set to start when the bladder tank reaches 22 psi and shut off when the tank reaches 28 psi, or some variation of those set points. That allows the pump to turn on when it needs to and run for a little while (long enough to run and cool) before shutting off again, whereas if there were no bladder tank, the pump would turn on every time someone opened a faucet for more than 5 seconds, the result of many such quick starts & stops would be high inrush currents, short operation times and perhaps overheating. With a large lake, you can probably get rid of the small tank, as its storage volume is more than likely redundant. I would advise against using it as a surge tank unless you determine its needed for such phenomenon, know its rated design pressure and ensure its safe for the maximum surge pressures you will have in this system. Otherwise, blind it off, as its just another potential leak.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com