Re: Archived Trends

[gorecki80]

Hi

I'm new in ZENON's world and I have a problem with Archive trend. My basic archive, that I need to show on screen is for a week. It's consisting of 404 variables.
Type of Scan is "On change"
Save is: saving format - Internal database (*.arx)
Saving cycle - 5 hours
Evacuation: Storage time - 2 years
Evacuation after storage time - internal database (*.arx)


My files are too large now - about 80 MB
before three days, we've had a problem(Power down for 7 seconds) . I tried to open trend, runtime doesnt draw, and after minutes Runtime showing error (image attached)
I have also and online trend and it is for 2 hours.

how is the right way to do arhived trend?

Thank you in advance

[markclemens]

Hi gorecki80,

Welcome to the forum!

It is hard to say from the screenshot what the cause for the error is. What is visible, is that you seem to be using sicam230, version 6.51. Also the ak_drv is involved in the error.

For sicam230 and especially the ak_drv driver, please contact your sicam230 siemens energy support. They should be able to assist you further.

In regard to the archive files, with spontaneous change and a spontaneous driver, the amount of values in a 5 hour archive can get quite large. (e.g. 404 variable values every 100ms = 4040*3600*5= 72720000 values (or 180000 values per variable for 5 hours)

When displaying this data in a trend screen that is 1024 pixel wide, this results in 175 values per pixel.

There are various approaches. First thing is to implement some form of deadbanding, either in the RTU or on the zenon side. Value changes in the 5th digit may not be significant, but can fill up an archive quite quickly.

Next is to tune your archives to what you want to view in the trend. Split your archive in several archives, having variables that you view together in a trend in the same archive.

Variables that you do not view simultaneously in one trend, in different archives. This way viewing the trend does not require reading an archive with 404 variables, but perhaps reading an archive with 20 variables.

Next is to tune the saving cycle. A shorter saving cycle results in smaller files. Save the basic archive e.g. every 15 minutes. Then implement following archives, saving e.g. every 5 hours. When viewing the past 5 hours, view the following archive, with e.g. 75 average, minimum and maximum values from the basic archive. When further detail is required, view e.g. the past 30 minutes from the basic archive.

best regards,
Mark

[gorecki80]

Hello, markclemens

Thank you very much for your immediate answer. I'll try the things you propose.
Can I make Online trend record for 7 days and how that will affect on Runtime?

Thank you very much