A firm in the US is designing systems for a process plant in China.  Over here, when we say "100 psi steam" it is understood that it means psig.  Over there, they use "bar" for pressure.  A 100 psig system would be 7 bar (gauge) or 8 bar (absolute).  In places where "bar" is used, is it assumed gauge or absolute?

If you go back to the original definition, a gauge calibrated in "bar" would show "1" at rest.  I saw one the onther day that showed zero at rest.  Katmar is right, never assume it is one or the other.  

I mostly work in low pressures and designing a compressor to go from 1 bar to 7 bar is a very different machine depending on what "1 bar" means (if it absolute then you're doing 7 ratios and need a two-stage machine, if it is gauge then you're doing 4 ratios and it is a pretty easy single stage).

Working in bars or psi is no different, bars are not some magical beast. Where there could be confusion I would specify "barg" or "bar(abs)" if I had to use those units, just as I used to say "psig" or "psia" if it was not clear one way or the other.

If you are talking design pressures for equipment, then you can be reasonably certain that specified values are in gauge pressures, but if for process design then it could be absolute. I find that Process Engineers quite often like to work in absolute pressures, and as zdas04 points out absolute pressures are a must in compression calculations (although I don't agree that a gauge at rest should show 1 bar, if at atmospheric pressure). It's taken a bit to train some Process guys that I want design pressures for vessels specified in "gauge" pressure.

But as katmar says, never assume anything. Someone once told me that to "assume" something means that you will probably end up making an "ass" out of "u" and "me". On occasion, I have been able to demonstrate that this statement is correct.

If you zero a gauge calibrated in bar(g) in Farmington, NM, USA then zero bar(a) would be -2.5 psia which is an impossible concept.

Pressure increments are arbitrary, but we have two physical points on the pressure scale that have historically been satisfied--zero absolute pressure is the minimum pressure that is physically possible, and zero gauge pressure is a convenient "constant" local reference point.  With other scales, both of these physical points are easily satisfied.  The increment on the bar scale requires that (except at a very specific elevation) they can't both be satisfied.  If I zero a bar(g) scale at local atmospheric pressure then I can't reach zero bar(a), if I calibrate the bar(g) scale so that 0 bar(a)=0 psia, then the needle will be somewhere other than zero on the scale at rest.

Since all of this is based on an arbitraty scale, I know that I can get to usable numbers by adding local atmospheric pressure to a gauge reading (in Farmington I would add 0.828 to the gauge reading to get to a universally usable absolute pressure), but looking at the corespondence I've gotten over the last couple of years, everyone who writes to ask me compressor questions just adds "1" to bar(g) to get to bar(a).  Maybe this is just a problem in Oil & Gas.  Maybe it is just a problem with us lazy compressor guys.  Maybe it is a universal laziness in the world.

My guess is that there are a huge number of bar(g) pressure instruments that are zeroed at local atmospheric and that the conversion from bar(g) to bar(a) is done by adding "1" to the bar(g) reading.  For most calculations, this is way the heck close enough.  For compression calculations from vacuum to a significant positive prssure it is a real problem.  For gas measurement calculations it would be a real problem.

I guess I spent too many years doing gas measurement to ever be cavilier about calibration and unit definitions.  I know that someone designing a relief valve for 300 bar(g) doesn't care if that is 301.000 bar(a) or 300.828 bar(a), especially since relief valve calculations are done in gauge terms.  The equations give you the same size orifice regardless.

If I'm designing a compressor to go from 103 torr (measured backward from zero gauge) to 3 bar (in Farmington compressing from 10 psia to 55.5 psia) if 0 bar(a)=0 psia then it is 5.5 compression ratios.  If 0 bar(g) = 0 psig and I assume that bar(g)+1 bar(a) = bar(a) then it is 4.0 compression ratios.  I would design the first as a two-stage recip (since I don't allow over 4 ratios per stage in the design step, you have to leave some slop for actual conditions), the second would appear to be an easy single-stage unit that would forever have problems with discharge temperature and/or rod load.

MORE NEWS