Can Venturi injectors be used with Flow Control Valves for maintaining ratio of mixing?

[kdv1988]

Hello guys,

I was wondering if you can control the flow-rate of the motive liquid & the suction liquid by using proportioning valves for both liquids to achieve the desired concentration?

For eg.. I have a venturi mixer, in which the motive liquid is water which is fed under gravity and the suction side is attached to a tank filled with coolant concentrate.

Now, with this if I have to achieve a concentration of say, 8% (knowing the volume of liquid to be dosed at that % concentration), I would calculate the volume of coolant to be mixed with water. Using flow sensors, can I control the flow-rate of the water line & coolant suction line by installing proportioning valves? or will restricting the diameter of the suction pipe affect the venturi suction? I am attaching a schematic for your reference.

For different concentrations, I would control the proportioning valves with the respective signals from my PLC. I am not looking to achieve critical levels of accuracy from this. An error of 0.5% in the concentration of mixed liquid should be acceptable.

Do you guys think this would work?

 

[EdStainless]

We used these for a coolant recirc system so coolant was our motive fluid. We had concentrated coolant as the side flow. The over all coolant level was controlled with a float switch and water makeup valve.
Coolant flow through the venturi ran all of the time for mixing. We had an inline sensor for concentration (I forget what it was actually measuring). When we need to add concentrate the side valve opened for a period of time (a few min) and then closed for at least 30 min. If we still needed more it cycled again.
At first we ran it with a needle valve in the concentrate line (and it's shut off valve) and whenever the water add valve opened so did the concentrate valve. But we couldn't keep this system balanced and make up for evaporation easily.

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

 

[LittleInch]

It might - difficult to say from afar. Not much detail in that block diagram.

Also what is the relative height / pressure in the venturi of the water and concentrate?

The issue is that your venturi doesn't generate a whole heap of negative pressure compared to the static head and hence there isn't much DP to play with.

IF the water flow is constant then you should be able to "trim" the coolant flow to get your 8%, but once you've got varying flows and pressure I don't think liquid/liquid venturis can do 0.5%

You might do better with a variable speed / flow metering pump or self powered injection pump where the motive power comes from the flow of the main liquid.

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

 

[kdv1988]

@LittleInch

Also what is the relative height / pressure in the venturi of the water and concentrate?
Water will be fed into the venturi under gravity from a height of approx 20ft. The concentrate will be stored in a small tank just above the venturi. So technically both are gravity flows.

IF the water flow is constant then you should be able to "trim" the coolant flow to get your 8%, but once you've got varying flows and pressure I don't think liquid/liquid venturis can do 0.5%
Understood. I cannot guarantee the flow of water to remain constant as the water inlet is basically a tapping from the plant's water supply piping. How about I have a small identical tank for water in my system? That way I can control the flow-rate of water. Instead of the service line directly feeding my venturi, the line would fill up a water tank in my system which would have a on/off valve. I would then allow this water to fall under gravity and act as the motive in my venturi. Not sure if there will be enough pressure in the line though. Height between the tank to the venturi would not be more than 3 feet.

You might do better with a variable speed / flow metering pump or self powered injection pump where the motive power comes from the flow of the main liquid.
Variable flow metering pump to dose the coolant into the stream of water right? Would that negate the use of venturi? How would the two streams mix in that case? I don't know much about the self powered injection pumps. I will read about those. Any particular reason you would suggest them?

Also, with metering pumps, the output is in the form of pulses right? Wouldn't that affect the mixing accuracy as well? Dampeners can be used, but I am just wondering.

 

[kdv1988]

@EdStainless

Don't think I quite understood what you described there sorry. Where would you add the water?

 

[EdStainless]

Into a holding tank. this was a central system that supported about 2 dozen machine centers.

For more consistent flow you could use a pressure regulator off of the water supply.

My point was though that you will have times that you just need to make up your 8% solution, times that you need to add just water to increase volume and bring it back down to 8% because of evaporation. So you need to have some independent control over these.

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

 

[kdv1988]

@EdStainless

I agree with the dosage of 'lean' coolant to adjust the concentration.

Somehow I am unable to fully visualize your system. You have a coolant line connecting 24 machines, that continuously has mixed coolant running through it. A concentration sensor monitors the coolant in this line.
When the need for an adjustment arises, you open the concentrate suction valve on the venturi to allow some concentrate to get mixed in the line. Wouldn't adding concentrate directly to a coolant make it even richer?
You balance that by adding water directly to the machine sump. Does that mean your coolant is mixed properly?
What happens when different sumps require different concentrations?

 

[kdv1988]

Also, I have another doubt not entirely related to this thread, but concerns mixing of two liquids. Not sure if I should start a separate thread for this, so if the mods feel that I should, I'll do the needful

Supposedly I have to mix coolant with water into a 20 Litre bucket at 3%. This translates to about 0.6 Litres of coolant mixed with 19.4 Litres of water.

If I divide the two flows into equal pulses, for eg. for every 5 Litres of water I mix around 0.15 Litre of coolant, but I do not control the 'flow-rate' of each, so for example if I start Line 1(coolant through a pump) and Line 2(water under gravity) together, the moment 0.15 Litre of coolant flows through, Line 1 stops till Line 2 has delivered 5 Litres and then Line 1 starts again. Thereby mixing 3% of coolant per 5 Litre four times over the course of 20 Litres. Would I still end up with 3% overall concentration? Technically I am adding 3% at the end of it. But the mixing of each 0.15 Litre of coolant is happening with not necessarily 5 Litres of water right?

And would this ensure a mixed coolant?

Inspiration -

https://www.youtube.com/watch?v=hFZgbdoZbSI&t=1245s

(From 19:30 onwards in this video)

Apologies if this sounds a little stupid. I just am not able to conclude.

 

[LittleInch]

First question first.

I don't know a whole heap about venturi mixers - you really need to look at and talk to some vendors to get actual flows, sizes, pressures, flow rates etc to see what is required.

8% mixing does sound like quite a lot, but maybe that's in the range of what they do.

The metering pump idea is an alternative to venturis and can give good fixed injection. You usually inject via a quill into the centre of the pipe flow.

Over any sort of decent flowrate the mixing will eliminate any minor pulsing of the injection fluid.

Second one I would flow the the coolant first which is what you're doing so I don't understand your bit about adding 3% at the end.

All depend son the flow velocity and what your bucket looks like. If there is a lot of swirl and turbulence in the bucket then it will mix well. If very slow then maybe not.

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

 

[kdv1988]

@LittleInch

Second one I would flow the the coolant first which is what you're doing so I don't understand your bit about adding 3% at the end.

That was just an example from the video to understand whether that method of mixing two liquids would yield a homogeneous mixture. Request you to just watch the linked video once from 19:30 in that video to understand my query.

 

[LittleInch]

I've seen that video before.

So long as there is a feature by which the concentrate is always flushed out of the hose or into the drum then you shouldn't get any loss of concentrate.

As said, the mixing is all down to turbulence and velocity of the tube end in the drum.

Or whether there is some form of secondary mixing in the drum.

ON the whole it is better to inject at 3% (or whatever) on a continuous basis into a flow and leave a sufficient distance in the pipe / tube from the injection point to the end point to mix properly - probably about 20 pipe diameters.

Venturi mixers are good but might be sensitive to the level of concentrate in your tank.

pump example

https://www.coolantconsultants.com/hydroblend™-6000-series-7-gpm-water-powered-proportioning-pumps

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