Mixed gas specific gravity calculation

[knapee]

Dear Sir

I would like to ask a very trivial question.
Though the last post

http://www.eng-tips.com/viewthread.cfm?qid=233225&page=20

Already talked a lot about this issue.
By definition the specific gravity of gas can be calculated by
(1). the (MW of the test gas)/(MW reference gas), or
(2). the ratio of gas density at actual conditions to air density at standard condition
I tried to calculate mixed stream (mostly is hydrogen gas) specific gravity.
Condition is in the following :
The mixed stream gas density =18 kg/m3.
The mixed stream gas’s molecule weight =36
Gas molecule weight = 29

By method (1) 18 kg/m3 divided the gas density (1.3) can’t be equal to
Method (2) calculation result: 36/29

Thank you very much for your help.

 

[katmar]

knapee, I suggest you restate your question more carefully. There are too many errors and inconsistencies in what you have written here.

You must be more rigorous about distinguishing between your flowing gas and air. You say "gas density (1.3)" and "Gas molecule weight = 29". These statements are confusing if the "gas" is not air.

Also if it is true that "The mixed stream gas's molecule weight = 36" then it is unlikely that your gas is mostly hydrogen.

You will see from the last line of my signature that I feel quite strongly about this.

Katmar Software - Engineering & Risk Analysis Software

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[hacksaw]

while it is a common practice in petroleum environments, to use the specific gravity of the gas, there are assumptions

The specific gravity of the gas(vapor) is the weight of the vapor at STP relative to the sampe volume of air also at STP. If the vapor contains condensibles, they are assumed to remain gaseous at stp.

It is also "common practice", though misleading, to use the term spec gravity at operating pressure, though it is really a measure of gas density. It is based on the "real gas laws" combined with an effective compressibility factor that is representative of the fluid. In that case the gravity is pressure dependent since the compressibility factor changes with pressure.

As a rule, usage is all over the map, so you have to do a little digging to find out the basis for the data you are working with.

 

[zdas04]

Hacksaw,
I have 5 references in my office that all agree that Specifc Gravity of a gas is defined as the molecular weight of the gas divided by the molecular weight of air. This is absolutely not density specific. A 0.65 SG gas will be 0.65 at atmospheric pressure, at STP, at 10,000 bar. It is an intrinsic property of the gas mixture.

For liquid, Specific Gravity is the density of the liquid at 60[°]F and 14.7 psia, divided by the density of water at the same conditions. This is not nearly as rigerous since all liquids are compressible to varying extents (liquid compressibility is the inverse of the liquid bulk modulus), but the error is often acceptably small.

David

 

[hacksaw]

David, no disagreement from this quarter, was attempting to explain some of the nonsense that takes place in operating (and less occasionally even in engineering environments). Admittedly in the last 15 years the data is much more rigorous and systematic.

In the past, I've been given "process data" for refinery projects, where the vapor data was given in the gas gravity at operating P&T and the molecular weight was only a number derived from lab measurments on sample bottles! Tricky stuff, and you had to get at the basis the guy was using in order to develop your own work. Most of the problem was one of communication, and easily sorted, but you had to be on the look out.

 

[jamesbanda]

All,

In regard to combining average density/ SG... for a system. It really depends on your system thermodynamics - operating pressures temperature on how to combine thermodynamic properties such as density. (and how accurate you need the data).. if you look into aspen your'll find many different ways to average properties..

 

[PaoloPemi]

I would presume that specific gravity is the density of gas at normal conditions divided by density of air at the same conditions, at "normal" condizions it is quite common to calculate gas and air density as Mw*P/R*t thus obtaining the specifc gravity of a gas as the molecular weight of the gas divided by the molecular weight of air. As alternative you can utilize a more accurate method (i.e. equation of state or equivalent) this is useful whith fluids (for example natural gases) where Z can be different from 1.0 , nowadays there are free software tools to calculate accurately gas densities and I prefer this approach.

 

[hacksaw]

The accepted methods are great if you have the data, the gas analysis, etc.

it is a bit of a problem if it is simply an "off-gas" or an "acid gas" of nominal properties...an all to common situation

 

[quark]

Using gas density to air density at STP for SG calculation is not wrong as long as we are not deviating from ideal gases(I remember we had discussed this issue earlier).

SG = MW_gas/MW_air
=> (MW_gas/24.05L)/(MW_air/24.05L)
=> (gram mole_gas/24.05L)/(gram mole_air/24.05L)
So, SG = Density of gas(kg/cu.mtr)/Density of air (kg/cu.mtr)at STP (by Avagadro's law)

 

[zdas04]

The whole "not deviating from ideal gases" thing bothers me with this discussion. If I calculate "Specific Gravity" as "density of gas / density of air" and disregard non-ideal behaviour, the "Specific Gravity" of something like Methane will be 20% higher at 1,000 psig than at STP. 20% is rarely trivial.

I'm not getting what the problem with stopping at the first step in Quark's parade of constants. Specific Gravity is a property of the fluid, not a property of the instantaneous thermodynamic conditions.

David

 

[25362]

IMHO, chemistry (and science in general) should provide a language that helps to turn everyday expressions into precise, unambiguous statements.

Since the definition of specific gravity rests on whether the density of both gases, the incognito and the reference, should be measured at equal prevailing T/P, or NTP, conditions, or only the reference gas should be at NTP (well defined) conditions, gaseous specific gravities are less in use, and densities at the prevailing conditions are preferred.

 

[zdas04]

25362,
The Specific Gravity of a gas does not rest on the density of the mixture at all. Specific gravity is the molecular weight of the gas divided by the molecular weight of air. No ambiguity. Specific gravity of a gas is a term used in thousands of fluid flow equations and wishihg it would go away is counter productive.

David

 

[dcasto]

Zadas,
I concur, but here is a monkey wrench in our simple world. The ISO standards update the Molecular Weight of air every two years.

If you want the details of calculating relative density, download GPA 2172.

 

[25362]

I prefer to quote katmar's observation in the post mentioned above by knapee, especially when working with gas mixtures:

"homayun, the MW ratio may not be logical, but it is the definition. I prefer to avoid all equations that are formulated around a vapor SG. It is much better to work with proper dimensions like mass density, compressibility etc."

 

[jamesbanda]

Surelly.

SG of a pure gas does not have any impact on mixture properties yes. But SG/Density of a mixture does have an impact on the properties.

 

[zdas04]

jamesbanda,
That made no sense at all.

The Specific Gas Constant is: R(gas) = R(universal)/MW(gas).

Specific Gravity is SG=MW(gas) / MW(air)

So, sinced you can calculate density using R(air) over SG of a gas instead of R(universal)/MW(gas) and get exactly the same number as in:

[ρ]=(P*SG)/(R(air)*T*Z)=(P*MW(gas))/(R(universal)*T*Z)

That says that SG/[ρ](gas)=1/[ρ](air) or the specific volume of air! Not a terribly useful result.


David

 

[jamesbanda]

David,

PV does not equal nRT for most gases.. when we are dealing with mixtures this formula can give cause for errors and more complex ways to average the formula are often needed for getting decent results. If your working at atomsphric pressure and temperature then you can use the ideal gas law without any risk but the further you deviate from the ideal gas state the more you have to use a thermdynamic equalition to model the behavhour.

E.g. when operating at say 100 barg, and high temperatures the ideal gas law would give you a poor model for the system.. and if you had a mixture of dissimilar gases then a simple average would not be approprate.. or if you had a system which formed dimers in the gas phase then equally this would not approprate.

 

[zdas04]

james,
If you use and EOS to calculate compressibility, then I've found that PV does in fact equal ZnRT well enough to match field conditions within the precision of the gauges I use. Shortcuts to calculate compressibility or leaving it out results in calculations that don't relate to measured data very well at all.

As I said above, SG/[ρ](gas) = 1/[ρ](air), which I still contend is not a terribly useful result.

I've found that if I have a gas mixture with over 95% straight-chain hydrocarbon gases, using SG allows most of the flow equations that I use to work pretty well. If I have more than 10% acid gas or aromatics then SG is a pretty bad way to extrapolate conditions.

David