jmw
Industrial
- Jun 27, 2001
- 7,435
See thread481-227786
I've progressed.
The latest version of the spreadsheet is Density 12MD.xls which is on my web site download page.
I have some assistance in cross checking it against the tables and so far we have found it can reproduce the tables.
It appears to be used by quite a number of people in the marine industry during bunkering.
This is a pressure operation - very little time, lots of things to do and watch.
Taking samples and analysing them is a problem, but with no online instrumentation, the way it is done is to use a hydrometer.
I got a fresh comment from someone who reported that the spreadsheet wasn't accurate.
He had a measured density of 740.6kg/m3 at 33 degC.
He gets a base density of 756.4 from 53B, he says, and
My spreadsheet gives 756.53kg/m3.
He has a program (Oil Survey) which gives 756.4
When I cross check with DNV PS I also get 756.4.
This doesn't look good.
I take a look at 53B.
The two columns are 739.0 and 741.0 with corresponding base densities of 754.9 and 756.9.
That's a problem.
By interpolating, I get a base density of 756.5 which is acceptable compared to the spreadsheet but why does my enquirer get 756.4 and in two independent programs?
I could do with some help here and if anyone knows the answer, please let me know.
Or, comment out this theory which I'm trying to check out:
This is the marine industry and it is bunkering where time is important and errors are to be avoided.
Some practices have evolved very little over the last 40-50 years.
I decided first to run a test of the tables and the spreadsheet at incremental values from 739.0 to 741.0.
Interesting.
There is a good correlation and simple rounding reduces the spreadsheet answers to the interpolated results from 53B.
But the DNV PS program and the Oil Survey program give a different set of results.
Any observed density from 740.3 to 740.7 gives the same answer - 756.4kg/m3.
53B was constructed by calculation with 0.25degC increments and 2.0kg/m3 increments.
We can interpolate density for greater resolution.
But interpolation means calculating while under pressure.
So I figure that what has happened is that for these applications someone decided to reduce the density increments to 0.5kg/m3.
Thus the operator would have a density table where the observed density of 740.6 would fit between two column headers of 740.5 and 741.0kg/m3. Since the point of this would be to avoid any interpolation during bunkering, it follows that the user instructions would be to round the observed density to the nearest value and simply report the tabulated result.
That means 740.6 is treated as 740.5 and the base density is then 756.4 - this explains all the apparent discrepancies.
Historically, the density tables were used as look-up tables. It is only recently that that the calculation has become the standard and not everyone uses calculations, mostly where there is online instrumentation.
It seems to me that if I'm right, there are some pretty reasonable justifications as to why they interpolate at 0.5kg/m3 intervals, not finer.
Of course, while the world has moved on, industry standards and normal practices are hard to replace.
I imagine the two programs must use look-up methods and the only way I can see this is if they store the expanded tables with the 0.5kg/m3 increments.
My only problem is that I need to find out if there is any foundation to this or is there a better simpler explanation?
Any ideas?
I have tabulated the data in the attached file.
JMW
I've progressed.
The latest version of the spreadsheet is Density 12MD.xls which is on my web site download page.
I have some assistance in cross checking it against the tables and so far we have found it can reproduce the tables.
It appears to be used by quite a number of people in the marine industry during bunkering.
This is a pressure operation - very little time, lots of things to do and watch.
Taking samples and analysing them is a problem, but with no online instrumentation, the way it is done is to use a hydrometer.
I got a fresh comment from someone who reported that the spreadsheet wasn't accurate.
He had a measured density of 740.6kg/m3 at 33 degC.
He gets a base density of 756.4 from 53B, he says, and
My spreadsheet gives 756.53kg/m3.
He has a program (Oil Survey) which gives 756.4
When I cross check with DNV PS I also get 756.4.
This doesn't look good.
I take a look at 53B.
The two columns are 739.0 and 741.0 with corresponding base densities of 754.9 and 756.9.
That's a problem.
By interpolating, I get a base density of 756.5 which is acceptable compared to the spreadsheet but why does my enquirer get 756.4 and in two independent programs?
I could do with some help here and if anyone knows the answer, please let me know.
Or, comment out this theory which I'm trying to check out:
This is the marine industry and it is bunkering where time is important and errors are to be avoided.
Some practices have evolved very little over the last 40-50 years.
I decided first to run a test of the tables and the spreadsheet at incremental values from 739.0 to 741.0.
Interesting.
There is a good correlation and simple rounding reduces the spreadsheet answers to the interpolated results from 53B.
But the DNV PS program and the Oil Survey program give a different set of results.
Any observed density from 740.3 to 740.7 gives the same answer - 756.4kg/m3.
53B was constructed by calculation with 0.25degC increments and 2.0kg/m3 increments.
We can interpolate density for greater resolution.
But interpolation means calculating while under pressure.
So I figure that what has happened is that for these applications someone decided to reduce the density increments to 0.5kg/m3.
Thus the operator would have a density table where the observed density of 740.6 would fit between two column headers of 740.5 and 741.0kg/m3. Since the point of this would be to avoid any interpolation during bunkering, it follows that the user instructions would be to round the observed density to the nearest value and simply report the tabulated result.
That means 740.6 is treated as 740.5 and the base density is then 756.4 - this explains all the apparent discrepancies.
Historically, the density tables were used as look-up tables. It is only recently that that the calculation has become the standard and not everyone uses calculations, mostly where there is online instrumentation.
It seems to me that if I'm right, there are some pretty reasonable justifications as to why they interpolate at 0.5kg/m3 intervals, not finer.
Of course, while the world has moved on, industry standards and normal practices are hard to replace.
I imagine the two programs must use look-up methods and the only way I can see this is if they store the expanded tables with the 0.5kg/m3 increments.
My only problem is that I need to find out if there is any foundation to this or is there a better simpler explanation?
Any ideas?
I have tabulated the data in the attached file.
JMW