In the current paper a model-based key performance indicator (KPI) is proposed to monitor and to suggest useful counter-actions, to be undertaken whenever inefficient management of cooling systems is detected, both at single telecommunication room and entire data-center or central office level. At the single room level, a simulator previously developed by the authors is used to simulate different cooling power trajectories as a function of actual and reference control strategy. The KPI values resulting from the comparison of the two simulations are then post-processed to evaluate significant statistical metrics, which in turn allow synthesizing the punctual information into a more comprehensive one. On the other hand, at the central office level the KPI index was calculated on a larger time-scale (i.e. up to 1 day) with respect to the single room case, by comparing currently measured power to the corresponding values simulated by a regression model specifically developed for such an application. The proposed methodology was tested both at room and central office level over an extended time window. The results obtained in the former case allowed to deeply analyze two alternative cooling management strategies, particularly addressing what is the most efficient one depending on outside temperature level. On the other hand, the KPI computed at the central office level was proven effective to verify if expected cooling efficiencies were met. Finally, the proposed KPI metric was discussed in view of its potential deployment to diagnose abnormal energy consumptions due to faults occurring either in sensors or cooling devices.

A Model-Based Key Performance Index for Monitoring and Diagnosis of Cooling Systems in Telecommunication Rooms and Data-Centers

SORRENTINO, MARCO;RIZZO, Gianfranco;
2013-01-01

Abstract

In the current paper a model-based key performance indicator (KPI) is proposed to monitor and to suggest useful counter-actions, to be undertaken whenever inefficient management of cooling systems is detected, both at single telecommunication room and entire data-center or central office level. At the single room level, a simulator previously developed by the authors is used to simulate different cooling power trajectories as a function of actual and reference control strategy. The KPI values resulting from the comparison of the two simulations are then post-processed to evaluate significant statistical metrics, which in turn allow synthesizing the punctual information into a more comprehensive one. On the other hand, at the central office level the KPI index was calculated on a larger time-scale (i.e. up to 1 day) with respect to the single room case, by comparing currently measured power to the corresponding values simulated by a regression model specifically developed for such an application. The proposed methodology was tested both at room and central office level over an extended time window. The results obtained in the former case allowed to deeply analyze two alternative cooling management strategies, particularly addressing what is the most efficient one depending on outside temperature level. On the other hand, the KPI computed at the central office level was proven effective to verify if expected cooling efficiencies were met. Finally, the proposed KPI metric was discussed in view of its potential deployment to diagnose abnormal energy consumptions due to faults occurring either in sensors or cooling devices.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11386/4254478
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? ND
social impact