measuring flow rate


Thread Starter


Hi Folks,

My dilemma is this, we manufacture pre-poured agar plates (around 30,000 a day) one particular part of the process involves inline mixing of liquid agar with horse blood. This is done via a Y connector. All of the pumping is done by individual parastaltic pumps and the tubing used is silicon (from 1/4 inch internal diameter to 3/4 inch). what we need to achieve is a mix of 5% blood to the agar. at present this is done by guess work, i.e. looking at the plates to see if they are red enough! What I would like to do is get this process more accurate perhaps by measuring the flow rate of the individual tubes and using that to work out a 5% mix. All of this takes place in a sterile environment.

Any thoughts or suggestions would be greatfully accepted.

Curt Wuollet

The parastaltic pumps should deliver a fairly constant volume. If you control and scale their speed you should have reasonable control of the mixture. Ideally you could use one motor with two rotors but 20:1 is a bit much to do with tubing sizes. I'd look for pumps that accept 0-10 VDC control signals and do as much as you could with tubing sizes, then scale the control voltage. I've seen multirotor pumps that accept different roller/tubing cassettes on the same shaft that are intended for this purpose.



Mark Hill - President, EESiFlo Inc.

Hi Michael,

Blood isn't an issue. We've measured it quite successfully in silicone tubes to 1/4", with accuracies well below 5%. Given that I don't know what consistency your agar is, it may be another issue altogether. If you'd like to discuss a solution to your problem employing non-invasive ultrasonics, give us a call.

Mark Hill - President, EESiFlo Inc.
mailto:[email protected]
Toll Free: 866-EESIFLO

Tomy Zacharia

Dear Micheal,

Any chance you can measure the RPM of the peristaltic pumps. These are Positive displacement pumps by the nature of their construction. From the dosage per rotation you can arrive at a rough figure of the blending percentage. The advantage being, you don't need to touch the process side.


Tomy Zacharia

Gerald Beaudoin

The situation you describe is very common to many industries where it is necessary to maintain the ratio of one component to another. Flow meters on the "wild" flow line and on the "injection" line both report to a single
ratio controller. The controller evaluates the ratio of the two flows as compared to the setpoint. The controller then provides a proportional output to control the amount of injected product. This is typically done via a VFD, which runs the injection pump. Using such a system, very close ratio tolerances can be maintained. The system simply follows any variations in the master flow rate, so you can start and stop, slow or speed up the line, yet the ratio always remains constant. ABB C300 series, among others, do a pretty good job at ratio control. I am sure there are others too. The PID blocks in some of the PLC's even offer this feature. Of course there are numerous vendors of flow meters and drives. Best to use what is locally available and popular. Good Luck.

Gerald Beaudoin

Michael Griffin

We've used positive displacement dispensers that were stepper motor and servo motor driven. While we didn't use them for mixing, they did give quite accurate results.

Perhaps you could use electronic gearing between the two systems. You could set the dispensing rate for one component, and have the other dispenser follow it.

Jeremy Ferrara

I feel your pain using parastaltic pumps. The real issues are that they quite often change flowrate to rpm ratio because of machanical failure over time in the tubing. Failure in the tubing is from many factors.

In fermentation, there is a method for feeding reactors using a mass demand. The feeding tanks are simply placed onto balances (to "weigh") and an electrical signal is sent to a computer from the balance; the computer then signal the pumps. So, you get a feedback loop on the pump speed related to the mass flowrate. This allows adjusting the pump speed to meet the ratios you need. Of course assuming you're getting compete mixing in your tubing.

One drawback is that the density of your liquids are going to change from batch to batch, but a given batch should have constant density given costant temperature and water content. You can track the densities and other qualities if you wish, however, each batch of product will be more similar than what you're using now. You'll get better batch homogeneity.

The cost of buying the balance equipment is going to the be the greatest expense, assuming you got someone onsite who can write a simple computer control program to run the pumps and respond to the mass changes over time. Since your entire process is dedicated to making these plates, it may be to your advantage to drop the bucks to buy the balances and cables (probable RS-232 or some variation.)

Hope this help!