Investigation of Proportional Integral Derivative control stability for mission critical cooling equipment

This paper details the areas of advanced research in control systems engineering and the practical application of Proportional Integral Derivative (PID) controllers for the temperature control of data storage units, which is a mission critical facility in the Royal bank of Scotland's data centre. Loop tuning techniques inclusive of the Ziegler Nichols were applied to obtain optimum system response, thus maximising efficiency of the cooling system. The business objective was to reduce the control systems offset to ensure the temperature set point within the ‘cold aisle’ containment remained within a ± 1°C error band. This ensured optimum climate conditions for the IT servers and optimised plant efficiency thus reduced the facilities energy demand. The control system comprised of a number of networked logic controllers, each with individual PID loop modules. Existing control system deficiencies highlighted the severe oscillations from inadequately tuned controllers; this led to uncontrolled cooling valves which ascertained a 5°C error band within the data storage ‘cold aisle’ containment. Application of temperature transducer calibration techniques, analytical equations, advanced tuning logarithms and software tuning provided a substantial improvement of stability control. Post Modification coefficients obtained an 18.4 % improvement with respect to decreasing the error band in all PID controllers. An expected lifecycle reduction of energy consumption from post modification tuning loops was calculated to be a value in excess of £238k.