by James Ottewill, Victor-Hugo Jaramillo-Velasquez and Michal Orkisz
Compared to other disciplines, the link between optimized condition monitoring and energy savings is somewhat intangible. Whilst an improved control algorithm or component design might reduce energy usage by some percentage relative to the existing state of the art, a calculation of the energy savings attributable to condition monitoring needs to consider a long period of the operational life of a component in order to capture maintenance actions associated with process stops. Nevertheless we may consider how energy savings might be made by properly establishing a condition monitoring approach.
Figure 1 gives a schematic of how the efficiency and the amount of energy required for maintaining a particular system might vary with time when no condition monitoring is employed. Initially after commissioning the efficiency of the system slightly rises as the equipment ‘wears in’ and operational best-practices for the particular system are established. After this point in time, the system performance begins to degrade due to aging. Limited maintenance actions are performed during this time, as the operator employs a run to failure approach. After a period of time, the degradation of components is such that the system fails suddenly and needs to be stopped for maintenance (unplanned stop). Because no monitoring has been employed, the energy required to correct the fault can be considerable as:
1. Necessary resources are not immediately available. Energy is required to ship resources as quickly as possible to remedy the problem; it is likely that, without proper planning, these logistics would not be conducted as efficiently as they could be.
2. Additional effort is required to fully establish the extent of the fault and where the root cause of the problem might be. This will likely require additional resources.
3. Whilst the root cause of the problem could have been a small, simple-to-replace component, the act of allowing the fault to propagate throughout the system might have led to the failure of a much larger component, requiring a greater amount of energy to manufacture and replace.
4. Points i-iii suggest that the unplanned stop in the unmonitored system will be longer and require more energy to correct. Throughout the outage period, there will be also be overhead energy losses (e.g. lighting, heating) which are not balanced by an output in product.
5. Because the stop is unplanned and in some sense uncontrolled, it is likely that there will be some loss or scrap of product with much of the energy previously input into the product being lost.
Let us contrast this to the case that a proper condition monitoring approach has been employed. Figure 2 gives a schematic of how the efficiency and the amount of energy required for maintaining a particular system might vary with time when condition monitoring is employed and a Condition Based Maintenance (CBM) approach is followed. Again there is an initial ‘wear-in’ period followed by some natural degradation of the system. Additionally, there is some energy required in initially engineering the condition monitoring solution, tailoring it for the system. After a period of initial degradation, the monitoring system might suggest that a particular incipient fault is developing, allowing a decision to be made on how best to perform maintenance actions before the failure propagates. A maintenance stop is planned, during which targeted maintenance actions are performed, see Figure 2.
Relative to a run to failure approach, a CBM approach offers the following improvements in energy efficiency:
1. As the stop is planned all necessary resources in place prior to the stop.
2. The root cause of the fault is indicated by the condition monitoring system, allowing maintenance personnel to target specific components.
3. The maintenance can be performed before the fault propagates to components whose faults may potentially require more energy to remedy.
4. With the stop planned and all resources in place to maintain the faulty component as highlighted by the condition monitoring system, the downtime is significantly reduced. The energy expended during the downtime is similarly reduced.
5. Similarly, with the stop planned, scrap losses can be minimized.
Additionally, the information from the monitoring system may also inform on weak points within the system which, if replaced, could improve the overall operation of the system. This will allow a planned stop focused on upgrading problem components and subsystems. Perhaps one of the other developments of the Energy SmartOps project might be incorporated during the upgrade, for example a more efficient control algorithm or a piece of turbomachinery might be maintained on the advice of advanced modelling predictions. Again the downtime is reduced relative to the unplanned stop, with the additional potential benefit of improving the efficiency of the overall system. The energy savings owing to the condition monitoring are related to improved operational efficiency of the system and reductions in energy required in maintenance actions, including reduced downtime during which energy is expended with lower quality, reduced or even no system output.
So potentially there are energy savings that are made possible by condition monitoring. However, there is an inherent difficulty in quantifying the savings (be they cost or energy) that are related to a properly implemented condition monitoring strategy. Specifically, and in contrast to our hypothetical examples given above, it is rare that there will be a true reference against which the success of the approach may be gauged. Consider a system which has a 70% probability of failing in a given period of time. Let us also consider that the system is augmented with a condition monitoring system which is used in the process of making operational decisions. Now let us imagine the case that after the period of time the system has not failed. Does this mean that the condition monitoring system has been successful or that the system did not fail because it is one of the 30% which do not fail? Ultimately, the energy savings attributable to a particular condition monitoring approach can only be distinguished after it has been applied to a large number of systems for a large period of time so that the law of large numbers is applicable and the statistics quantifying the success of the system can be properly extracted. When quantifying the impact of CBM on the energy efficiency of a plant, all of these considerations should be taken into account. Nevertheless, the potential energy savings offered by condition monitoring are evident and hold across a wide range of industry settings.