How differentiating between utilization of effective availability and utilization of effective capacity leads to a better understanding of performance metrics

In the recent past, the Performance Curve, or Operating Curve as it is sometimes referred to, has been used to compare the trade off between tool set utilization and normalized cycle time (x-factor). In most instances the independent variable for this comparison has been Utilization of Effective Availability. In this context, a tool is considered to be utilized whenever if it is actively processing product (described as being in production). For certain tool sets however, Utilization of Effective Availability is not a sufficient description of asset utilization. Batch tools are an example where actual utilization of the asset is driven by more than straight production time. Based on Utilization of Effective Availability, a batch tool is considered utilized whenever a batch is being processed regardless of the batch size. In this instance, a batch consisting of one lot utilizes the tool as much as a batch of four lots, Another example would be tools that can achieve a Parallel Factor greater than one. Parallel Factor is a way to measure the average number of batches being processed at the same time or in parallel. Photolithography clusters fall into this category since it is feasible to start another lot prior to the previous lot fully clearing the cluster. Again, the tool would be considered utilized if any part of the cluster was still processing product even if other parts were being starved. For both example tool sets, normalized cycle time (x-factor) is lower than expected when plotted against Utilization of Effective Availability on the Performance Curve. For tool sets such as these, a more refined measure of utilization which takes into account effective capacity and not just availability should be considered. This paper describes some of the underlying variables that determine Utilization of Effective Capacity and provides case studies to demonstrate the effects on performance metrics.