Mass flow meter and orifice type flow meter choose dilemma 2

[hswang2]

We have a steam stream with orifice type FT max flow rate : 7000kg/hr min flow rate :500 kg/hr but no normal flow rate given in the datasheet. I&C staff told us the orifice type FT (due to the orifice narrow range usually 10:1 for given transmitter span setting, apparently in our case already exceed 10:1 span range so he suggest us change to Mass type FT). But it seems not necessary to use Mass type FT in steam service. Or could we stacked the conventional FT in parallel to cover the wider range or any other better solution to solve this high turndown dilemma in this case.

Thank you very much for your kindly help.

 

[JLSeagull]

The ration of 7000kg/hr maximum and 500 kg/hr minimum is FAR GREATER than 10:1. With dp type flow meters that transmitt the orifice differential the relationship between flow and dp is exponential. The flow rate is not a correct comparison for a linear mass flow meter output and a dp orifice flow meter. Compare the 4-20 mA signal output for turndown differences. Depending upon the steam temperature, other technologies exist such as vortex shedders etc.

 

[dcasto]

You can use stacked DPs and get 20 to 1 turndown with an orifice. Tripple stacked can get 100 to 1.

An orifice meter is a mass meter, we just set them up to read volume.

 

[ChEMatt]

I am not familiar with the option of stacking orifice meters. hwsang2 uses the term "parallel", so in my mind I'm seeing two or more orifice meters in parallel, with similar branched piping (perhaps not necessary to be symmetrical?)

If this is the case, is hswang2 going to have to cut into the pipe to weld in the new meter and associated piping? If that's the case, would it be better to go with another kind of meter and replace the existing orifice meter?

Onwards,

Matt

 

[dcasto]

Stacked DP is where you have just one primary element and you place 2 or 3 DP transmitters on the tap across the orifice.

One DP transmitter is ranged 0 to 10" of water, the second is 0 to 100" The first has a range of squareroot of 5% of the 10" range or .5 to 10, or .7 to 3.16. The second has a range of squareroot of 5% of 100 or 5 to 100 or 2.23 to 10. The net flow range is .7 to 10 of maximum flow or 14 to 1 turndown. Add another transmitter ranged 0 to 250 " and you can get 25 to 1 ratio.

 

[ChEMatt]

I understand, dcasto. Thank you.

Onwards,

Matt

 

[hswang2]

Dear
sorry to bring this up agian. Because the previous discussion, the max/min flow rate ratio is much greater than 10:1 so one orfice flow meter can't cover the range.

I put the two orifice type
flow elemnt in series on the P&ID to replace one orifce
flow element. But one dilemma occur in my mind, the high range type FT put in front of the low range FT, what if the differental pressure on the small orifce is way too large?

Thank you for your help.

 

[zdas04]

You misunderstood DCasto's solution to your problem. In short, he's saying that you size your (single) orifice plate for a very low dP at minimum flow and a very high dP at maximum flow. Then put a pair of pressure transducers on the dP port (both "runs" can share the Static transducer).

The key to this working is the fact that modern transducers have an uncertainty based on their calibrated range. So if you have a +/-0.5% uncertainty, when you calibrate the transducer to 0-50 inH2O then the uncertainty is +/- 0.25 inH2O, if you calibrate it 0-1000 inH2O then it is +/-5 inH2O.

In this scenario one of the transducers measures a very low dP with great sensitivity (say it is ranged for 0-50 inH2O) and will give you numbers for the low-flow case. The other dP transducer is calibrated to a very high range and gives you flow when the other one over-ranges (most of the time). The low range "run" measures from 500 to probably 2000 kg/hr (turndown is 4) and the other one measures from around 1500 to 7000 kg/hr (turndown is 5).

It is a really elegant solution that I've never come across before. I'm not sure exactly how you program an RTU to know which one to ignore, but that is "just" programming (definition of "easy" is "someone else has to do it").

David

 

[StewPad]

Hi guys I think you are missing something.
it's all depend on how much dP can you spare on the orifice.
If you have a large margin you can do both reading in one orifice
Send me your pipe and orifice spec and I will calculated it back

 

[dcasto]

all flow totalizers I've ever worked with have a call function with a name like "PICKBESTDP". The function recognizes stacked DP's and chooses when to switch between them. You could do the same with a DCS/PLC with a simple if then else routine.

 

[rocketscientist]

Have any of you guys considered a multivariable dP, such as a 3095 MV. The span is larger than the 4:1 you can expect for a standard, uncompensated orifice. In addition, steam is a wet gas. Orifice accuracy declines with a wet gas without compensation --- use a MV.

I like dcasta's idea except for the error from different transmitters. This is always a problem.

I think a single MV with a 7:1 turndown might do the trick. Otherwise, size an MV to handle the min-nominal flow and do as dcasta suggests but size an orifice for the largest flowrate at 10-12:1 turndown. Maybe an MV for the larger flowrate would be a good idea too.

 

[dcasto]

us old schoolers keep it simple, and you are correct. We had a station where the flow kept jumping between the two transmitters during certain runs. We notice a larger meter error. We re spanned the two transmitters to keep from jumping between the two transmitters and our meter error went away.

In the old Daniel 2500, it had a call function called "BESTDP(DP1,DP2,DP3)"

 

[davefitz]

When using an orifice flow element with wide flow turndown ( stacked DP) one needs to be careful regarding zero setpoint erros due to condensation/ bubbles/ air contamination of the sensing line- one can show a finite DP when in fact the DP is caused by the different density fluid in the sensing line. Also, without contamination, a long vertical sensing line can have a different density fluid due to thermal differences between process temp and the sensing line temp.