• Not Answered

Calculation of TSS on DeltaV Insight Models Viewing.

Hi everyone,

I would like to know how the TSS calculation is made in DeltaV Insight which is reported in Models Viewing section.

I know that TSS calculation, i.e. for FOPDT process response should be calculated as: (5 x Time Constant) + DT

However, for one of my control loops DeltaV Insight reports the following figures in Models Viewing:

a.  TC= 356.2 sec,  DT= 91.8 sec and TSS= 1055 sec

Doing the manual calculation, I obtain TSS= 1872.8 sec  =  (5 x 356.2) + 91.8

b.  TC= 316.6 sec,  DT= 94.1 sec and TSS= 960 sec

Doing the manual calculation, I obtain TSS= 1677.1 sec  =  (5 x 316.6) + 94.1

Could any one help me, in order to understand how DV Insight makes TSS calculations and why is so much difference in comparison with manual calculations?

Best Regards.

• TC
• TSS
• DT

Marcos Moreno

• 11 Aug 2016 5:20 PM
• Cancel

3 Replies

• lsocorro

  

  • 11 Aug 2016 5:31 PM

  it should be multiplied by 3 based on this "Typical estimates are: (3 × First Order Time Constant + Dead Time) or (4 × Dead Time), whichever is greater for the slowest responding input/output pair"

  • Cancel
• Marcos Moreno

  

  • 12 Aug 2016 4:00 PM

  In reply to lsocorro:

  Hi Isocorro,

  Yes, you are right, that is what BOL explained, but if you do the math with the data reported in DeltaV Insight, the calculations do not match!

  Example, data taken from DV Insight Models Viewing tab for one of my PID Loops:

  				3 × First Order Time Constant + Dead Time	4 × Dead Time
  TC	DT	TSS	% Difference	TSS-1	TSS-2
  342.2	95.1	1055	6.32%	1121.7	380.4
  316.6	94.1	960	8.74%	1043.9	376.4
  281	102.5	1148	-17.64%	945.5	410
  315.6	102.5	1148	-8.60%	1049.3	410
  342.7	98	1342	-16.09%	1126.1	392

  As you can see, if one does the math there is a difference between DV Insight and the calculations done with the indicated criteria. I included a column labeled % Difference, you can see that the difference is big.

  So, the same question remains, How does DV Insight do the calculations for TSS?

  Thanks.

  Best Regards,

  Marcos.

  • Cancel
• Mark Coughran

  

  • 15 Aug 2016 6:51 PM

  In reply to Marcos Moreno:

  This thread seems to be getting DeltaV Predict confused with DeltaV Insight. The reply from Isocorro gives the definition of TSS for a multivariable problem from the BOL topic "Using DeltaV Predict Pro", whereas Marcos asked about DeltaV Insight which applies to SISO problems.
  
  I did not find in BOL a formula cited for TSS in DeltaV Insight. However, if there is a statement, using the same formula (3xTime Constant + 1XDead Time), then I suggest you have already proved it from your table. An agreement within 10% or even 20% is the order of consistency you should expect from Insight or any other process identification tool. For example, if you are measuring Insight process models with a nominal output of 40%, the process gain is likely to vary significantly when you move to 60%. So my answer is another question: what do you intend to do with TSS?

  • Cancel