Natural Gas Metering Options 3

[fizzle]

Hi Guys,

I'm looking for some direction in picking meters to monitor Natural Gas flow and consumption.

I have a number of large box fires operating at max fire rate of 8 million Btus/hr, and am trying to decide on a technology to meter the amount of gas they use in an hour.

I'd like to stay in the realm of Mass Flow, if possible, but cost could make me move to volumetric calculation.

Does anyone have any experience to share with monitoring Natural Gas with Coriolis or Thermal Mass Flowmeters? Thermal Mass meters sounded promising, but I'm told they are subject to large amounts of drift over time.

Ideally, I'd like a meter that requires little maintenance, and is easy to calibrate.

For reference, the pipe size I'm working with is 2" sched. 40, with Natural Gas at 18psi. Turndown is a consideration as well, as my furnaces can go to low fire of around .5 million Btu/hr.

Any experiences/recommendations?

-Matt!

 

[quark]

I have seen orifices and annubars with Rosemount's multivariable dp transmitters (3095MV) working perfectly well within 1% error but turndown can't be morethan 1:9(theoretical). These meters give you ease of maintenance and calibration(you just have to inject dp signal).

For higher turndowns, I prefer corriolis massflow meters. Calibration for these meters can't be done at site easily.

Regards,

 

[jay165]

How 'bout an ultrasonic:

http://www.daniel.com

 

[fizzle]

I'll check into them! Thanks!

 

[TD2K]

I 'think' Corolis meters are starting to be used for gas but I don't think they are down in the 18 psi range.

Orifice plates are pretty common in this type of service. Ultrasonics are another option. I don't have any experience with thermal meters to comment.

 

[phex]

In our most recent projects, we specified coriolis flowmeters in all fuel gas applications. I think this could be comparable to nat gas service, but we always had about 50 to 60 psi.

Just my two cents.
chris

 

[friartuck]

What volumes are we talking about...small or large. (kW duty or m3/hr will do)

Friar Tuck of Sherwood

 

[phex]

In my case we're talking about 3300 Nm3/hr through a 3" coriolis meter.

 

[friartuck]

I know this might sound like a daft question, but if the medium you are measuring is gaseous, why do you want to measure the mass flow rate?

Calorific values for gas are normally stated in volumetric units and therefore doesn't it make more sense to measure that value?
Incidentally, we use turbine meters which have a very low pressure drop and have built in digital outputs which we send to BMS (Building Management Systems).



Friar Tuck of Sherwood

 

[quark]

Mass flow meters also give you flow measurement in volumetric units. The main advantage is that the volumetric flow is corrected by density compensation(which will, otherwise, give you error in reading due to temperature and pressure changes of flowing units). The meaning of volumetric and mass flows shouldn't be taken literally.

This article deals with natural gas flow measurement by dp and turbine meters.

http://www.flowmeterdirectory.com/flowmeter_artc/flowmeter_artc_04031102.html

Regards,

 

[bogdanm]

I'm working in the Natural gas metering field,and I can tell you in this way: the pressure that you have it is to low to use a Coriolis meter or an ultrasonic meter. At this pressure,I suggest you to use a gas turbine with a flow computer (S600 from Daniel is the best I've ever see), tipically a gas turbine has 1:20 turndown and error is 0,5%.The flow coputer can gives you the volumetric, the mass and the energy flow. Keep in your mind that for an accurate mass and enegy flow you must use and a gas chromatograph, or to ask the gas supplier about the gas composition.

 

[hacksaw]

bogdanm's comments reflects my own experience.

without sufficient the fluid density the vibrating or vortex shedding methods don't work.

keep in mind that the turbine types do require protection against overspeed and need provison for oiling on line (if the bearings arn't sealed.

 

[jsummerfield]

Both coriolis and thermal mass are well suited to the application.

I regard any head type meter as limited to about 5:1 even with today's accuracy. The burners too have limited turndown. I would not object to orifice plate technologies at this size.

What precision is required - and why?



John

 

[khan101]

I have seen multi-path ultrasonic meter in custody transfer applications. Krohne is one of the manufacturers who makes 3-pass ultrasonic, good enough for custody transfer. The only advantage I see though is no pressure loss. Rosemount multivariable is also a good choice but calibration stability is a proble. Coriolis is generally limited to 6" line size. Coriolis or turbine would be my choice as well if pressure loss and cost is not an issue.

 

[fizzle]

Guys,

To answer your questions:

-Volumes are 9000 scfh or so at max fire, 500 scfh or so at min fire.

-Volumetric meters require pressure & temp compensation for calculations, mass flow meters do not. I'd prefer to not backwards engineer the BTU content per scfh if possible. Really, this is just a way to remove some error if I can.

-Precision: I'd like +/- 1.5% total error, if possible. Mainly, the meters will be used to prove out process improvements and to try to troubleshoot why one furnace may be running differently than another.

-Matt!

 

[Cockroach]

I've designed and used orifice and venturi meters in this application, ANSI Class 150. They are enormously low cost compared to others mentioned, particularly ultrasonics which are very poor below NPS 6 line size applications.

These days a guy could strap a Rosemount or equivalent dP meter to the side for telemetry out to a surface datum logger or signal processing unit. Either way, that would proove far more costly than orifice/venturi meter itself!

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[steveen]

Fizzle,

Check out V-cones. Very accurate and repeatable.

http://www.mccrometer.com/products/pdvcone.htm

They are less expensive, but do require corrections for pressure and temperature.

They also do not require extensive upstream and downstream piping runs.

On my installation, they saved a bundle due to the piping configuration.

 

[jsummerfield]

This is not a V-cone application. Piping upstream/downstream is not an issue at 2" schedule 40. Also, in my experience the V-cone is not consistent with the +/- 1.5% total error accuracy.

BTW, I saw a new element at ISA with a venturi and annubar that has the benefits of a V-cone and better accuracy. I have no experience but am interested in this technology.

John

 

[Bubbarock]

I would have to agree with cockroach and would go with orifice style in this case. 2" sch 40 simplex or even OFU if you are not changing the plate at all are the best bets and with a cost of approx. $800.00 per unit well below any others.