Traditionally, commissioning is that point in new construction or a major retrofit when, with the project’s systems design engineers still on hand, building operators verify the performance of all systems. Retro-commissioning refers to doing it all again in one year, five years, ten years — whatever the contracted frequency. Too often, commissioning is a ‘hot potato’ job that, while all-important, is tossed around due to lack of budget and lack of a champion with the right mix of IT and building automation skills. The alternative of Continuous Optimization was conceived as a way around these issues: If controllers could share data about their states such that equipment could tell you when performance begins to lag, buildings could almost fix themselves. In 1999, buildings scientist Tom Hartman describes this potential solution:
If you are a designer who wants to improve the comfort and environmental quality of your projects and at the same time improve their energy efficiency, it’s time to consider network-based control. By employing network-based control, equipment can operate at improved efficiency and also operate to meet the loads served more effectively. Network control is possible because most DDC systems now have the ability to “network” (or share the value and status of points) among controllers.
In February 2005, two building scientists from the Pacific Northwest National Laboratory (PNNL) introduced the term ‘Continuous Commissioning’ when describing how web services – along with IT methods of gathering and analyzing big data sets – can be put to use improving building energy and comfort performance. Here’s an excerpt:
How will commissioning likely be performed 20 or 30 years from now? With the push of a “start” button, equipment and systems should all test themselves, identify any installation or configuration problems, automatically fix problems amenable to “soft” solutions, and report the need for “hard” solutions requiring replacement or installation of hardware. A report on the performance of all building systems and equipment should be automatically generated, delivered to key recipients, and stored electronically for future reference and updating. During initial operation (e.g., for the first year) and continuing over its lifetime, the system should optimize itself, integrating its behavior with external constraints, such as occupancy, occupant behavior and feedback, energy prices, demand charges, and weather.