Sizing a booster pump for gas chlorination

[undertan1995]

I'm seeking some advice for something that should be pretty simple. I have a tendency to over-complicate, but here goes:

Background info:
I've got a project in which a water treatment plant receives water at an average rate of 1 MGD, with a max rate of 3 MGD. After standard chlorination math I've found that the chlorination system will need to be able to feed a maximum of 125 PPD. However, the plant currently has a chlorinator rated for 250 PPD. For conservative reasons I've decided to ensure that the chlorine ejector has enough supply pressure and flow to output 250 PPD, even though it really only seems like 125 PPD is sufficient.

The system has a booster pump which withdraws a water supply from the incoming water main (conveying 1-3 MGD), boosts across the ejector, and injects in the water supply main at a point downstream from the withdrawal point. I have determined that the pressure at the injection point is approximately 46 ft (20 psi). The 1" solution line from the ejector assembly to the injection point is approximately 85' with several bends and fittings.

Problem:
It's my understanding that the backpressure seen on the ejector depends on the system flow (injection point pressure + headloss from injection point to ejector). Obviously you can't calculate the headloss without knowing the system flow. So that leads me to selecting a pump. The chlorinator companies don't really dabble in booster pumps. They just say you need a particular flow and head, and I have yet to find an example online where this is gone over in detail.

I plotted the system head curve looking from the pump discharge to the injection point. I assumed that the static head condition over the pump is essentially zero (technically slightly negative), since it's taking water from a pipe under pressure and discharging it right back into that pipe. So the system head is literally controlled by friction losses. So I started plugging in flows and heads along my developed system curve to see which would be me the proper combination of ejector backpressure and supply pressure/flow. Since the system head curve will only show me the head instantaneously at the booster pump, I had to factor in flow and headlosses to the ejector inlet.

My question is, am I overcomplicating this? Or does this seem like a reasonable way to design?

 

[bimr]

The 20 psi at the injection point seems to be low as that is the minimum pressure for distribution systems.

You need a minimum pressure of (20 psi + 3 psi + piping headloss) at the injection point. Then, select the injector and add the required injector pressure to that of the injection point.

Here is a link with an example:

example

 

[LittleInch]

I don't understand why you're doing it this way.

If you've got to put in a pump why are not simply putting in a dosing pump?

If you do go this way, the best option is to use some sort of PD pump which gives you good control on flow and the pressure just takes care of itself.



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

 

[bimr]

This method is common practice for handling chlorine gas. Chlorine gas is pulled from a pressurised cylinder under vacuum through supply lines. This is the safest way to feed gas chlorine, any break in the gas line results in a loss of vacuum instead of a dangerous pressurized gas leak.

 

[LittleInch]

Ok, I forgot it was gas...



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

 

[undertan1995]

Thanks for your replies. I guess I'm getting hung up on the whole pressure drop and backpressure concept. I've attached a schematic. In this case the only way I can reason is that the pump sees a headloss of 87-30= 57 psi at the ejector at its operating flow of 17 gpm. If it did not, the flow would have to be insanely high to lose 87-20 = 67 psi from the ejector to the injection point. How else would a pressure gauge read 30 psi just downstream of the ejector while another gauge reads 87 psi upstream of the ejector?

 

[bimr]

The injector performance is from the manufacturer. Yes there is a large pressure drop across the ejector.

250 lbs per day chlorine = 4.7 kg/hr.

Using the chart for EJ-5000 Ejector. 30 psi (2.07 bar) = backpressure. The chart gives a 4.5 bar (65psi) inlet pressure and 105 lpm (27.7 gpm) flow.

You are taking water from the pipe at 20 psi. (your page shows 55 psi?). The pump boosts the pressure from 20 psi to 95 psi (20 + 65 + 10). The ejector has s pressure drop of 65 psi. That leaves 30 psi on the discharge side of the ejector.

I assume you have 7 psi head loss in the discharge piping after the ejector. That would leave you with the pressure in the bypass line of 23 psi which is then discharging into the main line which has a pressure of 20 psi.

Confirm with the manufacturer that the EJ-5000 Ejector is the correct ejector size for 250 lbs/day.

 

[SteveWag]

Sorry about being so late. With vacuum "ejectors" it is not unusual to see 3 to 5 times the input pressure as the output. Be conservative about headloss in small piping, sometimes there are many more fittings that you expect. I have used mostly turbine pumps if I felt the calcs were trustworthy. If you are using a centrifugal pump, be generous with flow and head and make SURE the motor is non-overloading over the entire curve.
Steve