Natural Gas flow with ML1400

[bigbadtimemachine]

All,

I have a project that has stumped me.....
This is an upgrade from a Bristol Babcock flow computer to a new ML1400. This NG gas site has two new DP meters scaled 0-600 inch of water column in 4-20ma range, one temp inlet sensor, one inlet pressure, and HMI that inserts orifice plate hole size.
I have to get from this setup the MCFH rate, then store TODAY, YESTERDAY, MONTH, AND LAST MONTH.
My problem is I am unsure how to convert inwc to MCFH using the AGA8 standards. I have never done a gas application
Any help would be great !

 

[Tom Jenkins]

The mass flow rate is proportional to the square root of the inches water column differential pressure and the density. The density is proportional to the absolute temperature and absolute pressure of the gas. The math looks intimidating, but is actually fairly straight forward.

I suggest you get the equation from the orifice plate or DP transmitter supplier. Many times the transmitter does the square root extraction for you, but if your reading is inches water this may not be the case.

If the supplier can't give you the equations then go to either AGA8 or ASME PTC 19.5 texts.

 

[bigbadtimemachine]

Tom, thank you for the quick response!
Ill call rosemount and see what they say.

 

[diat150]

IMHO the right tool for this job is a flow computer.

 

[Tom Jenkins]

IMHO the right tool for this job is a flow computer.

Why?

 

[diat150]

Because the flow computer will do everything that he has asked for, and has already met all approvals for measurement of natural gas. Not only that, it will be easier to calibrate, pull audits, and historical data that is required in natural gas measurement for custody transfer.

Im not saying you cant do it in a plc, but in my experience, you are are much better off using the tool that was designed for the job.

 

[Tom Jenkins]

If you are talking custody transfer you certainly have a stronger case for the flow computer, Diat, as it adds an element of confidence for the purchaser. For control and normal data logging purposes though, all of the calculations and functions needed can be done in a PLC - I've done so. (Of course, this presumes the programmer understands the math and physics of the flow measuring system.)

 

[grnick50]

In the past I succesfully used thermal mass flow meters for biogas from a company called Combimass.
http://www.bindergroup.info/binder/en/products/flow/gasflowmeasure.php.

 

[bigbadtimemachine]

This is one of those projects that was sold before the Engineer knew what he was getting into, so now I'm stuck with doing the dirty work. This data is used to keep the gas supply company honest not for a billing application so its not 100%. I have no way of knowing the makeup of the gas at a given time so I'm using DP, Orifice plate hole size, temp,static pressure to come up with CFH for a IFIX SCADA display and report.

 

[Tom Jenkins]

This site can get you started:

http://www.engineeringtoolbox.com/orifice-nozzle-venturi-d_590.html

You will definitely need pressure and temperature measurement to correct the density. Natural gas is methane - you will need to include its properties in your calculations.

 

[diat150]

This is one of those projects that was sold before the Engineer knew what he was getting into, so now I'm stuck with doing the dirty work. This data is used to keep the gas supply company honest not for a billing application so its not 100%. I have no way of knowing the makeup of the gas at a given time so I'm using DP, Orifice plate hole size, temp,static pressure to come up with CFH for a IFIX SCADA display and report.

Id definitely check with the customer and make sure they know what is going on. for a true aga8 calculation, you will need to use quite a few constants like methane, gravity, btu, etc to properly calculate the flow. You can get a flow computer for a couple grand. By the time you implement this logic, and the fact that the micrologix isnt an approved measurement device, it may make more sense to just get the flow computer. If this meter is a "check" meter, then the company that you would make an argument against will not recognize the data that you provide. you will also not have audit logs to show that your meter was properly calibrated and a historical data trail that cant be manipulated by an outside source like in the micrologix.

Just a thought, dont shoot the messenger. Ive had to deal with these types of installations before and it is almost always a disaster and ends up getting replaced with a flow computer...

 

[bigbadtimemachine]

Thanks,
What I have so far is the 4-20ma is scaled to 0-600 inch water
Today my inch of water is scaled is a value of 238, sqrt 15.427248620, temp is around 81.9
static pressure of 113.1
One DP meter is off line and the other is flowing with a orifice plate with a 2.000 hole in it.
I can come up with most of what i need but the orifice plates are changed based what gas company is flowing so the HMI will input data for hole size....and then I'm lost.
The ML1400 has compute command that I'm using to get CFH but the hole plate size has me stumped as to get the right numbers. I called Emerson and spoke to a tech and he was helpful but not a PLC guy.

 

[Mickey]

Thanks,

The ML1400 has compute command that I'm using to get CFH but the hole plate size has me stumped as to get the right numbers. PLC

Can you show us the "CPT" instruction or the math you are using to get your results?

Do you want "Standard" cubic feet or "Actual" cubic feet?

Changing orifice size changes the coefficient (CPrime)used in a flow calculation.

 

[bigbadtimemachine]

Mick,
thats what has me stuck . As to how and where to use temp and pressure I have no clue ?
I've never done a NG site so this is all French to me.
I keep being told that a flow computer is only way to do this, but the old system that was in place only had DP, temp, psi, and plate size. So it can be done.

 

[Tom Jenkins]

The orifice diameter or the "beta ratio" should be stamped on the orifice OD or a nameplate on the orifice plate.

The general equation of the mass flow rate through an orifice is:

q_m=n π/4 d^2 Cε√((2ρ(ΔP) g_c)/(1-β^4 ))


The symbols are defined in ASME PTC 19.5, along with better representations of the equation and a lot of additional equations.

The flow can be calculated in a PLC - I've done it, and in the days before floating point and compute blocks to boot. However, note my caution above - "this presumes the programmer understands the math and physics". If you don't understand Bernoulli's Law, Boyle's Law, and Charles' Law then you will either have to learn them, work with someone that does, or go to a flow computer as Diat suggests.

This is another example of the truism that writing the code is the easy part - understanding the system and the physics behind it is the tough part.