Triad is a Federal/State Interagency Partnership
Scheduling and Load Balancing
Successful Triad projects require careful attention to scheduling and load-balancing issues.
Dynamic work strategies and real-time data collection are two of the three primary Triad components. The purpose of both is to facilitate efficient and effective decision-making. The presumption, however, is that data can be collected and reviewed quickly enough to support real-time decision-making and keep activities on track. The bottom line is that under the Triad, there is an additional level of choreography and coordination for field activities that is not required for a more traditional, static set of work strategies. This becomes of critical concern during remediation, when the implications of work slow-downs or inefficiencies in field activities can have significant repercussions for project budgets and schedules. Triad case studies have demonstrated that coordinating real-time data-driven field activities can be done in a manner that does not impose an undue logistical burden.
One of the questions that must be decided is the frequency of decision-making in response to real-time measurements. For example, suppose the purpose of data collection is to delineate areas of known contamination via sampling and real-time analytics. Additional sampling location decisions will be based on prior results. One way to approach this is to truly make sampling sequential, with each new sampling location predicated on the results of the previous locations. Alternatively, one might collect batches of samples consisting of more than one location, and then determine the locations of the next batch based on the previous batches' results. The extreme in the latter case is the traditional static work plan where all sample locations are predetermined, and follow-on sampling is postponed to another field effort.
Determining the optimal size of data collection "blocks" is one of the design requirements of a Triad-based data collection program. There are a number of factors that will affect this design. They include sample turn-around times, sample throughput rates, the nature of project decision points (e.g., timing, sequence, level of project management or core team involvement, time requirements for conflict resolution, etc.), the amount of data required to support specific decisions, and the ability to group data collection work into discrete sets of activities performed sequentially. For example, if the turn-around time for a sample result is a minimum of 24 hours, basing the next sampling location selection on the prior sample's results will not be possible. If the throughput of a field-deployed analytical laboratory is a dozen samples per day, then a dozen samples might constitute a logical block of data upon which to build a program. If soil excavation work is organized around remediation units, and a minimum of five samples are required to draw decisions about each unit, then five samples (or some multiple of five for multiple remediation units) might be the appropriate organizing framework for scheduling. Finally, if a program integrates characterization with closure activities, a strategy might be to sample likely contamination areas on the first round, return to bound contamination where found on the second round, and then return once again to perform closure sampling outside the boundaries of contaminated areas for the final round. In this case the sequence of required activities is the basis for scheduling and coordinating sampling and decision-making.
The concept of load balancing, which is intimately related to scheduling, also is important to the Triad. The concept is simple: matching analytical capacity to sample production and throughput in a manner that optimizes the overall characterization or remediation program. Optimization here means achieving confident decision-making at lowest overall project costs. The importance of load balancing lies in the fact that analytical over-capacity (i.e., under-utilized analytical capabilities) is a waste of project resources (since for most real-time measurement systems, per sample cost is a function of sample throughput), while analytical under-capacity can result in project delays, or time critical decisions that are made without the requisite supporting data.
There are several ways to promote effective Triad scheduling and load balancing. These include:
- Rotating data collection/analysis activities and decision-making among different locations at a site so that there is sufficient time to produce and review results before a subsequent decision needs to be made for any specific location. Activity rotation provides a mechanism for field teams to move on to another location to continue work while waiting for results from the previous location and associated data collection activities.
- Staggering work assignments within a work day to support 24 hour decision turn-around. For example, the data collection team might work a 7 a.m. to 4 p.m. shift, the analytical team the noon to 9 p.m. shift, and a decision support/data review group the 3 p.m. to midnight shift, with subsequent work decisions ready for the next morning. This requires the availability of decision-makers at unusual hours, and so would not be effective for decision-making that involves Triad core team members.
- Providing for "overflow" analytical capacity that can be used if required. One example of overflow analytical capacity is having a fixed laboratory available that can provide fast-turn around (albeit at higher per sample expense) if field deployed equipment cannot keep up with work flows. Another example would be to have additional real time method instrumentation readily available to be used as necessary.
- Cross-training technical field staff so that personnel can be re-assigned to tasks associated with activities that become bottlenecks. For example, on-site analytical service providers often have chemists that have been trained in proper sampling techniques and can be utilized to supplement the sampling team during slow on-site laboratory throughput. This type of technical staff cross-function responsibility-sharing will require upfront planning and may affect subcontract terms and agreements.