Commissioning, including acceptance testing, is a somewhat rigid process intended to verify and validate that a project delivers what is required and expected. As with most processes it is paramount that each step, phase, or activity is begun and completed sequentially to keep the process flowing smoothly. At a high level, this is obvious. You shouldn’t start the design phase until the programming and requirements definition phase is complete and you shouldn’t start construction until the design is complete, etc.
This strategy holds true for acceptance testing as well. Acceptance testing should progress from factory testing, to equipment receipt and installation and continue through startup, followed by performance testing, and culminate in integrated systems testing (IST). Furthermore, the testing should start with verifying the simple before proceeding to the complex. Start with verifying components, then equipment followed by systems and finally culminating with IST.
There are two basic philosophies in play. The first is to identify and resolve problems and discrepancies as early as possible so they can be resolved with the least impact to schedules, budgets, and quality. It is easier to correct manufacturing errors prior to shipment, to address damaged equipment upon receipt rather than during installation, and to resolve installation errors prior to proceeding with performance testing.
Likewise, performance testing should proceed from verifying simple components before testing assembled equipment, and testing equipment before proceeding to systems testing, and not attempting to test system interfaces until all respective systems have been validated. When commissioning is executed consistent with these strategies, each activity builds increasing confidence that the ensuing activities will be successful and keep the identified discrepancies to a manageable level.
The converse is also true. When commissioning and/or acceptance testing fails to follow these simple strategies, the result is often adverse impacts on schedule, budgets, and quality. Unfortunately, this happens far too often even on projects where the participants are seasoned veterans of major construction projects where formal commissioning is the norm.
What causes the commissioning process to get compromised? There is no one answer. In many instances, it is a combination of competing influences and poor assumptions that result in decisions to eliminate steps or to attempt to schedule actions that should be sequential into concurrent/parallel paths. Some examples include assuming manufacturers have sufficient internal quality controls and testing to preclude shipping products with manufacturing errors and defects so factory witness testing is eliminated to save time and money. Another questionable assumption is that shipping and receiving staff will identify and correct damaged equipment upon receipt, or that reputable contractors and vendors will install and startup equipment thoroughly. The reality is even the best products on the market have occasional defects and the best contractors make mistakes, especially when they are pressured to meet schedule no matter what.
Most construction projects have detailed master schedules that lay out a somewhat ideal timeline for when tasks and activities should occur. These schedules are “living documents” that tend to evolve as the project progresses to reflect the real-world realities such as late shipments and deliveries, adverse-weather related construction delays, coordination issues amongst various trades, etc. Most schedulers include some “contingency time” to allow for minor delays, rework, and other unexpected delays
Common sense would find that as scheduled activities slip, the overall project timeline should also slip. But in many cases the project team begins trying to “compress” the schedule rather than delay project completion. The first things to go are the “contingency” times. After that the project team begins looking for what activities can be performed in parallel vs. in-series. This often results in incomplete equipment startup and checkout with equipment and systems being declared ready for functional testing when in fact they are not ready and in some instances, aren’t even safe for operation.
It is not that people don’t want to do the right thing and deliver a complete and quality product. But as quality starts to conflict with schedules it becomes somewhat subjective as to what “ready” means to various entities. The mechanical contractor may consider the installation of an air-handling unit complete before the electrician has completed wiring up the unit, and the electrician considers the unit ready for operation even though the controls contractor hasn’t had an opportunity to setup the controls. And then comes the debate about whether the units are ready for performance testing before the test and balance contractor has had the opportunity to balance the systems.
Some common startup issues that commissioning engineers encounter all too frequently include incomplete incorporation of monitoring and controls, factory defaults that do not match site specific conditions, loose wire terminations and connections, incorrect breaker settings, incomplete water and air TAB (test, adjust and balance), etc. In many of these situations, the project team presses the commissioning agent to forgo the requirement that all startup and checkout activities be completed and to proceed with performance testing and “assume” the contractors did quality installs and the equipment and systems are ready. The reality is that what is being decided is to combine level-3 startup and checkout with level-4 performance testing which often becomes a recipe for disaster.
In worst case examples, failure to perform thorough equipment startups results in equipment damage and possibly staff injuries. Protective devices that are not set properly or tested can result in shorts, over pressurizations, overloads, and other destructive failures. Any schedule improvement that was hoped to be gained is now lost and then some, not to mention the added costs associated with repairs or replacements, rework, and lost time as mobilized staff wait for the rework to be completed.
In my experience, trying to skip or otherwise compromise the startup and checkout phase inevitably results in many more problems, delays, and quality issues than would have occurred if the schedule was adjusted to allow for a complete and comprehensive startup prior to beginning performance testing. It is like trying to expedite cooking a meal without allowing the oven to finish preheating. You can increase the cooking temperature to try and rush the process, but eventually the meal turns out poorly.