Sizing Dosing Pump for 93% Sulfuric Acid

Hi,
answer 1)
Note: for your application consider Milton Roy pump (chemical dosing pump)
Pierre

1 - Yes, confirmed

Note that depending on the pump design the actual pulsing flow may be 2-3 times higher than that average.

Hi shvet and Mr. Pierre,

Thanks for your quick replies. I will prepare my specification approach Milton Roy Pump for quote today.

shvet said:

Note that depending on the pump design the actual pulsing flow may be 2-3 times higher than that average.

Click to expand...

SO if I need 0.5 gpm, I will actually get 3*0.5 = 1.5 gpm flow?. So the stroke length has to be adjusted to get the required flow rate then?.

Thanks and Regards,
Pavan

OP,
Unlike a centrifugal pump, your flow rate selected is an average flow rate based on the cyclic displacement of the pump not a constant flow rate. When applying the Darcy Weisbach equation, keep this in mind that your flow rate may not be laminar, that said, there is a direct correlation between cycles, stoke length and velocity that you can use to determine the maximum Reynolds number and friction head. Some pumps use a diaphragm type accumulator to help alleviate the pulsing, but it should be recommended, upon commissioning, that a down-stream flowmeter be installed to confirm the required flow and not just rely on the pump settings.

OP,
Just a general note of caution. When designing system or specifying properties be aware of where you are getting your information. A manufacture's pamphlet may be helpful, but it also may not be giving you the whole story. Always try to start with industry standards and best practices such as NACE and your ChemE bible, Perry's. Understand, the manufacture most times is not publishing their own information but is pulling from other sources, which may be out of date or has changed. Also, never except a manufacturer's claim out right without a) Previous experience with that manufacture for that application or b) Requiring the manufacture to "show their work" so that you can empirically verify their claims.

Hi Heaviside1925

Heaviside1925 said:

OP,
Unlike a centrifugal pump, your flow rate selected is an average flow rate based on the cyclic displacement of the pump not a constant flow rate. When applying the Darcy Weisbach equation, keep this in mind that your flow rate may not be laminar, that said, there is a direct correlation between cycles, stoke length and velocity that you can use to determine the maximum Reynolds number and friction head. Some pumps use a diaphragm type accumulator to help alleviate the pulsing, but it should be recommended, upon commissioning, that a down-stream flowmeter be installed to confirm the required flow and not just rely on the pump settings.

Click to expand...

The flow rate that I used to calculated the frictional head is 0.5 gpm which I want. Are you saying that I should use the maximum flow which is 3 times the required flow and use this to calculate the TDH?. Yes I will suggest the client to use to Pulsation Dampener with N2 gas as the dampening medium. Also will suggest them to use a Flow Meter to meter the flow rate.

Heaviside1925 said:

Just a general note of caution. When designing system or specifying properties be aware of where you are getting your information. A manufacture's pamphlet may be helpful, but it also may not be giving you the whole story. Always try to start with industry standards and best practices such as NACE and your ChemE bible, Perry's. Understand, the manufacture most times is not publishing their own information but is pulling from other sources, which may be out of date or has changed. Also, never except a manufacturer's claim out right without a) Previous experience with that manufacture for that application or b) Requiring the manufacture to "show their work" so that you can empirically verify their claims.

Click to expand...

I got the density and visosity from Narfalco Sulfuric acid manual. I think it is a reliable one. For the vapor pressure I could not obtain data from Perry. All I could find in Perry's is the partial pressure of water vapor over aqueous sulfuric acid solution. As Sulfuric acid solution in water is not ideal, I cannot calculate Sulfuric acid's vapor pressure using Raoult's law. Any suggestions on how I can calculate it?. Yes I will ask the manufacturer to show their to verify thatb they are using the correct properties.

Thanks and Regards,
Pavan Kumar

OP,
If you are using a pulsation dampener, then using the average flow rate should be adequate for your application. The point I was trying to make and what I am assuming shvet point was the same, is that you need to understand the kinetics of the pump you are choosing and keep that in mind when calculating additional parameters. In other words, do not treat it as a "black box" with a certain flow rate coming out but understand the mechanics of how that energy was imparted into the fluid flow.

OP,
In regard to your second question.
To clarify, my second post was only meant as a note of caution. I have had experience with engineers using vendors pamphlets as an engineering basis and ending up with a poor design. In your case, its sounds as if you have done your due diligence and I meant no disrespect.

Hi Heaviside1925,

Thank you for valuable inputs. I did not get offended or anything of that sort. I value your comments highly.

Now if let's say I don't use a pulsation dampener then I have to use the maximum flow rate for my head loss calculations?. And do I have to correct the pump curves due to high viscosity of Sulfuric Acid(25 cP)?.

Thanks and Regards,
Pavan Kumar

OP,
Are you saying that I should use the maximum flow which is 3 times the required flow and use this to calculate the TDH?. No, TDH for a metering pump should be provided by the manufacture and may be quite high in the case of a metering pump.

To give you an example of the concept I was trying to communicate. Let's look at a simple piston pump rated at .5 gpm. The piston has two strokes, one to fill, one to discharge. Even though the average rate over a one minute period is .5gpm, only half that time (30 seconds) is fluid being discharged. If you have a stroke length 6" and you are at a set rate of 60 cycles per minute. Then the max fluid velocity would be 60ft/min (60 discharge strokes every 30 seconds).

This is a very over simplified example, and your downstream piping design should be based on the actual characteristics of the pump you select. I was only communicating this concept for your consideration when looking at the overall dynamics of your system.

OP,
To your last post.
Now if let's say I don't use a pulsation dampener then I have to use the maximum flow rate for my head loss calculations?. And do I have to correct the pump curves due to high viscosity of Sulfuric Acid(25 cP)?.

pierreick suggested Milton Roy. They are an excellent company and will provide you with like support. Metering of H2SO4 is a very common application. They should be able to supply you with all the information you need. There are other manufactures, and I would inquire as to where else at your facility metering pumps are being used and what manufacture was selected. It's important to also provide a pump that your maintenance and operations folks are familiar with and that they and you can solicit local vendor assistance for.

Pavan Kumar
I also wanted to thank you. Many times, on this forum, people posting are trying to short cut doing the actual engineering work themselves or just don't understand the engineering concepts they are trying to apply. You are the exception and I have enjoyed giving you what knowledge I have. You did impart some to me as well. I have sized pumping applications for non-Newtonian fluids before, but I have never had an application to considerer non-Newtonian dynamics in respect to metering pumps, which would be different than other pumping applications. I now will, thank you for another tool in my toolbox.

Hi Heaviside1925,

Heaviside1925 said:

OP,
Are you saying that I should use the maximum flow which is 3 times the required flow and use this to calculate the TDH?. No, TDH for a metering pump should be provided by the manufacture and may be quite high in the case of a metering pump.

Click to expand...

I need to fill my specification and send to the vendor and in the spec I specify the required flow rate and the Head to be developed by the pump based on the suction and discharge head losses. I did not quite understand why the manufacturer would give the TDH. I think they will just give the curves. Will these curves have to be corrected for viscosity as 93% Sulfuric acid has 25 cP viscosity.

Heavyside1925 said:

To give you an example of the concept I was trying to communicate. Let's look at a simple piston pump rated at .5 gpm. The piston has two strokes, one to fill, one to discharge. Even though the average rate over a one minute period is .5gpm, only half that time (30 seconds) is fluid being discharged. If you have a stroke length 6" and you are at a set rate of 60 cycles per minute. Then the max fluid velocity would be 60ft/min (60 discharge strokes every 30 seconds).

This is a very over simplified example, and your downstream piping design should be based on the actual characteristics of the pump you select. I was only communicating this concept for your consideration when looking at the overall dynamics of your system.

Click to expand...

Please let me know if I understood it correctly. With a velocity of 60ft/min based on a stroke length of 6" and 60 cycles/min, to get 0.5 gpm flow out of the pump the discharge port needs to of Area, A = 0.5*0.133681/60 = 0.001114 ft2 with diameter D = SQRT [4*0.001114/PI()] = 0.037 ft =0.45 inches.

Now when the stroke length and number of cycle/min is set and with the given discharge port size the flow is fixed correct?. How would the flow increase more than this. Please excuse me if I did not understand the concept correctly. I would then use 0.5 gpm for my head loss calculations correct?.

Thanks and Regards,
Pavan Kumar

OP,
Will these curves have to be corrected for viscosity as 93% Sulfuric acid has 25 cP viscosity. The manufacture will supply.

TDH for a piston pump is simply force over area. The force will be the motors max torque it can apply to the piston. This can be much higher that one is used to seeing on a centrifugal pump. Since this is a function of force applied to the piston, the manufacture will need to supply it.

Concerning the last question. What I was showing was not meant for calculation purposes. If friction head is determined by the average flow rate only, it negates the max velocity apparent from each "pulse" of the piston. This may or may not be important, but I only mentioned it to keep in mind that the actual friction head may be higher than what you calculate using only average flow rate. My only correction in your calculation, as a concept, is that you are assuming that each cycle is equal to a certain volumetric displacement, which it is not, it is equal to an average volumetric displacement. The actual volumetric displacement being imparted with velocity only exists for half the cycle, the discharge portion. As noted, this is specific to the pump along with the fluid properties. It could be only a third of the cycle is used to discharge and the remaining 2/3rds is for the piston to fill. This is where your higher velocity pulses exist and create more friction head than just based on an average flow rate over the same delta T. To not belabor things, a pulsation dampener/accumulator with negate most of this. Your vendor support should be able to answer any additional questions.

Hi,
My advice, don't over specify, let the vendor doing his job.
Other vendors can supply this technology (Prominent, Pulsafeeder,Etc)
What about piping material?
Good luck.
Pierre