Chemical Injection Flowmeters and Control Challenges 1

[Sawsan311]

Dear All,

Based on your field and design experience, for a system having multiple injection points fed from one single head injection pump, there are mainly two challenges experienced:

- flowmeters not able to measure or detect the low flow rates below 0.3 l/hr. What would be an optimum flowmeter to be selected for such a readibility issue.
- the control of the injection points might mandate installing Internal recycle control valves IRCD.. any experience on this technology?

What are other possible technologies available for mitigating the low flow rate readibility as well as the finetuning control.

Multihead pump design is the last resort since it will require changing existing design. A brownfield solution is much more needed.

Regards,

 

[LittleInch]

I'm not quite following your system design or your question.

Is this basically a single pressurised header and several offtakes?

How is injection controlled?

0.3 litres per hour is incredibly low.

You might be able to utilse fixed orfifices if your injection pressure is high compared to the fluid being injected into, but no data.

Look up laboratory metering systems for that level of flow.

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

 

[TiCl4]

I've injected critical raws into a process at a large dP (100 psig down to 20-50 psig) at flow rates as low as 0.3 gph, no problem. I'm assuming that your process requires various injection flow rates, given you are talking about flow control valves.

For flow measurement, I've used low-flow Coriolis meters. Micro-motion LF Coriolis by Emerson or Promass A 300 from Endress Hauser are both low-flow meters that read in your required range. Be aware that small coriolis meters are susceptible to vibrational interference - don't use these in high vibration areas.

I used a low flow control valve by Badger Meter to control flow. They have a low-flow RCVs (research control valve) with 100% open Cvs ranging from 0.0000018 to 6.0, which can control well within your lower range. I've pasted the link below to the RCV trim datasheet so you can peruse it. Note that the Cv values provided are for 100% on the valve. You'll have to call the manufacturer to get specific Cv values below trim code JEP - the Cv vs %lift chart available online only goes to "J" trim.

Adding individual control valves and flowmeters is your alternative to multi-head pumps or separate pumps. For this application (from the little data you've provided anyways), I recommend your main supply pump (centrifugal?) be designed to recycle back through an orifice into the supply tank in order to control flow and provide back-pressure for injection.

If the application is very flow critical, I suggest you ensure the flow through the recycle line is at 5-10X or more (my rough guestimation) the maximum flow rate of all your injection points combined. This is done in order to keep the pump at relatively the same operating point and keep a relatively constant header pressure no matter what the injection demand is. If you size your pump/recirculation flow too small, the drop in pressure from another injection point ramping up would cause temporary drops in flow at other points until the control valves opened.


Link

https://www.badgermeter.com/resourc...ication application data rcv-as-00582-en.pdf/

 

[mk3223]

The issue could be what's the actual accuracy of 0.3l/hr flow rate for your measurement as required by the process. And then, you may select the proper elements to adjust and control the operation.
Without a multihued pump system for each stream design, you may need to spend some time (or a lot of time) to calibrate one pump- multiple branches to accomplish the flow accuracy as needed.

 

[Sawsan311]

Thank you TiCl4 and everyone for your response..

Indeed, the system involves more than 10 injection points with ranging capacities and having 0.3 l/hr approx as numerous points capacity being very low for measurement and controlling..

 

[TiCl4]

Note that the Coriolis meters generally have an accuracy of +/- 0.1% of measured value, so they are very accurate. You aren't likely to get much better accuracy on any other instrument measuring down that low. That, combined with a properly sized control valve, should meter your solution accurately with complete ability to measure in your flow range.

To repeat a question from LI, do you currently have an existing header (single pipe) with multiple injection branches coming off of it? How do you meter the injection now? Whats the viscosity of the pumped material? Can you provide a sketch of the current system so we may make suggestions for the easiest/simplest retrofit to meet your needs? I gave my prior suggests by assuming a great deal about your current design, but it may not be the easiest retrofit, depending on your current design.

 

[Sawsan311]

Thanks TiCl4, actually the system is simple, single head pump feeding more than 10 injection points from a main supply header like a fishbone. Each injection point is having a rotameter which is quite poor in measuring commonly.

The flow rate can reach very low value at some points i.e 0.02 l/hr as a chemical dose. Density is between 750-1000 Kg/m3.

 

[TiCl4]

I will give you a word of caution. If you do pursue this, be aware that the orifice size on these control valves if you are controlling at 0.02 l/hr somewhere around 0.003 inches (Cv near 0.000018 or so). That's ~75 microns. You will need to ensure your fluid is EXTREMELY clean, likely putting a 25-um absolute filter upstream of the control valve, to ensure the valve does not become plugged.

If the solution contains a dissolved salt close to its solubility limit, be aware that large a dP across the control valve can cause flashing of liquid and and increase of localized concentration of the salt. This can lead to salt precipitation and deposits on the valve orifice that will choke it off. I found that particular gem out the hard way.

Lastly, 0.02 L/hr is much different from 0.3 L/hr. The smallest Promass A 300 I mentioned has a max flow reading of 0.735 lb/min, with a max turndown to 0.001 lb/min, or ~0.027 L/hr at 1000 kg/m3. You would be pushing the limits of commercially available industrial equipment.

The only other option I see to pursue, though, is to look at syringe pumps for lab applications. They can meter down to a sub-uL/min range. However, that would require a complete change for you to go to a multi-pump system.