Public Content Library

Symposium: ETS Operational Efficiency Through Automation and Machine Learning 

12-11-2018 02:48


Use of Statistical Process Control for the Plant, Network and ETS, Veerendran Krishnan 
Statistical Process Control measures and tools are used to improve the performance and to provide better service to the customers. Standard Deviations, Control Charts, Harmonic Means and Multi-variable linear Regression are some of the measures used in this perspective. The measures are implemented in the Control System consisting of PLCs (Programmable Logic Controllers) and SCADA (Supervisory Control And Data Acquisition) Systems. The process has thrown a lot of light in to the performance of various flow control arrangements, Quality of Service claims, effectiveness of the implemented control philosophy.
The Standard deviations are mainly useful wherein we try to quantify the level of deviation from the intended path or value. An average will not reflect the true picture of the deviation since the negative errors may nullify the positive errors. The use of SD is particularly useful in the following scenarios,
a. To provide a measure for the District Cooling plant chilled water supply temperature’s adherence to the set/committed values. Here the SD is calculated against the set/committed chilled water supply temperature. The resultant SD is a good measure for the customer/operator to know where it stands in terms of the supply temperature set point.
b. To provide a measure for the effectiveness of the Temperature Control valve in the ETS. The valve is modulated in such a way that a specific return water temperature is achieved or a specific secondary side supply temperature is achieved. The SD which is measured with the actual temperature against the set temperature provides a measure for the effectiveness of the implemented logic/PICV.
c. Use of Multi variable linear regression help us predict a dependent variable such as ETS Header Return water temperature against the other ETS independent Process variables such as ETS Header supply temperature, Flow rate, HEX secondary supply temperature, HEX Return temperature and Valve open positions. Such predictions will help find any anomalies that may arise on account of sensor failures
d. A similar use of Multi Variable Linear regression will help us find the Plant Supply temperature (dependent variable) when we know the flow rate in each chiller along with its leaving water temperature and de-coupler flow rate along with its temperature (all independent variables).

Plant Operational Audit & Optimization, Jad Honeine
BB-03, Empower’s newest DCP, has all its major auxiliary equipment operating on advanced energy saving machineries. An energy audit was conducted to detect daily accumulated wasted KW consumption. This consumption was continuous and was never expected even when the equipment was not running. Empower’s team cooperated and proposed the Best Operational-Methodology to either minimize or neglect this loss. The case study proposed accommodates all types of plants and can be implemented with no financial aid.

ETS Automatic Verification Tests, Taha Fahim
The presentation is based on a technique to invoke a series of automatic sequences that would enable operator to record the system responses and study any anomalies. ETS Heat Exchangers (HEX) are subjected to full flow conditions while the flow in the other Heat-Exchangers are shut-off and instrument readings are recorded after a delay. A logical model is applied on the read data and the deviations are flagged for further verification and rectification.

ETS Temperature Control with Cascade and Feed Forward Control, Shamma Bin Lahej
The presentation describes using a cascade-control/feed-forward control methods, with flow rate as Process-Variable for second loop, to respond to flow changes for ETS, even when temperatures remain at desired set-values. The changes in the secondary side supply temperatures indicate the upcoming changes in the primary side temperatures with a time delay thus secondary side temperature changes through control valve can regulate flow even when the primary temperatures are not affected resulting in better temperature management.


Veerendran Krishnan, Empower
Jad Honeine, Empower
Taha Fahim, Empower
Shamma Bin Lahej, Empower


0 Favorited
4 Files
pdf file
4B.1_BinLahej.pdf   916 KB   1 version
Uploaded - 12-11-2018
pdf file
4B.1_Fahim.pdf   310 KB   1 version
Uploaded - 12-11-2018
pdf file
4B.1_Honeine.pdf   1.35 MB   1 version
Uploaded - 12-11-2018
pdf file
4B.1_Krishnan.pdf   1.35 MB   1 version
Uploaded - 12-11-2018