Justification for data historian?

[controlnovice]

What is the justification for a data historian in, or connected through OPC, a DCS or PLC system?

Application of the control system is control of a small chemical plant ~3000 I/O.

______________________________________________________________________________
This is normally the space where people post something insightful.

 

[JLSeagull]

The data permits analysis of the causes of an upset or a trip. Perhaps long term data collection could prove useful if a state or federal agency asks about your emissions, releases, accidents or other operating issues on July 21, 2007.

 

[IRstuff]

Intermittents, and previously undiscovered errors, would be my justification, with the latter being more critical.

Often, you'll find a "new" problem, but careful review of past data will show that the problem existed for a long time, but was only now actually "discovered."

The historical data might then show the conditions what result in the "new" problem, and save you months of investigation.

TTFN

FAQ731-376

 

[ScottyUK]

The trend and event recording capabilities of a good Historian are invaluable when trying to investigate a complex event such as a process upset with no obvious cause. They are useful to provide evidence supporting things like warranty claims, performance tests, and the like. A well configured Historian removes a lot of speculation and guesswork, and can be a good tool for keeping an eye on how well operational staff are performing. They soon realise that the computer just records what happens impartially so things like shift logs become more accurate!


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[roydm]

I work with a bunch of process engineers, they seem to love crunching historical data in Excel. I suspect the Data Historian is better at saving in that format.
Roy

 

[dcasto]

downside is, you have a smoking gun. Make sure you have a written proceedure for getting rid of old data.

 

[itsmoked]

Also they often create a tremendous pile of data to warehouse.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[IRstuff]

True, but at least, it's in digital form.

TTFN

FAQ731-376

 

[PNachtwey]

The data can be used for modeling systems and tuning PIDs.

 

[danw2]

This strip chart recorder ad from the '40's is just as valid today as ot was then:

 

[ScottyUK]

That must be the grandfather of the ones I work with!


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[roydm]

What's a strip chart recorder?


Oh - I remember that's one of those things we bought years ago for chewing up paper isn't it?

Roy

 

[JLSeagull]

As I recall, Brown became Minneapolis Honeywell. There were lots of Brown recorders, especially multi-point temperature recorders in the old refineries when I was a pup.

In the early '80's we put data loggers in plants that had two paper boxes. One would feed the printer, the second to collect the printed paper. Nobody at the plant evaluated any of the data - only moved the paper to storage. I used the records to determine the time between a field gathering compressor trip and the plant inlet pressure falling to the low pressure alarm point, etc.

 

[dcasto]

I'll tack on to JLS's post. Even with a data historian I make the operators keep some paper logs just they do not sit around. Their are always sites and sounds that real senses can detect. Just a quirk.

I also like paper chart back ups on sales meters, the lastrst flow computers fail and a paper chart comes in handy when there are $100,000,000 at stake.