P2-3. AHU and VAV Box Diagnostics > Conclusions

Conclusions:

• Both APAR and VPACC are suitable for embedding in commercial control products at little or no added direct cost. Many existing control products already have the computational horsepower needed to use the embedded rules. The initial cost to manufacturers will be in engineering and field evaluation to assure reliability. The APAR and VPACC rules, or variations on them, are highly likely to available commercially within the next five years.

Commercialization potential or commercialization initiated:

Several commercial partners provided control products for use in the Virtual Cybernetic Building Testbed. The FDD rules were successfully embedded in the controls products by programming the APAR and VPACC algorithms in the native language of each product. There are four major buildings controls manufacturers that have expressed interest in working with NIST to embed the diagnostics in their respective products. These features may become the basis for optional features that increase revenue, or standard features that are used to distinguish a product from the competition.

Recommendations:

Full commercialization of automated fault detection and diagnostics is one in which APAR and VPACC, along with appropriate parameters and thresholds, are packaged within HVAC control products. To become reality, more work is needed in three main areas.

Built-in APAR and VPACC functions would greatly simplify the task of embedding FDD in a control program. It is impractical to expect trend data to be evaluated to determine the necessary parameters and thresholds for each site, as was done in this study. Ideally, sets of robust parameters and thresholds that are effective across specified ranges of applications would be available. Additional field data from a wide variety of systems must be collected in order to determine these robust parameters and thresholds. Also, the current embedded FDD tools are written using generic mathematical functions available in the languages in which the controllers are programmed.

More work is needed to develop alternative ways to interpret FDD results and deliver this information to the building operator. For example, rather than automatically sending the alarm to the operator, the building control system could highlight, on demand, those devices having experienced the greatest number of alarms in a given period of time. Or, if an automated maintenance management system is used, an alarm could automatically generate an appropriate work order. However, many faults are the result of design or commissioning issues that are beyond the scope of the building maintenance staff.

A mechanism is needed to resolve multiple conflicting fault reports before reporting them to the operator. A fault in another piece of equipment, such as an air handling unit, boiler, or chiller, could result in a large number of alarms, perhaps overwhelming the operator.

Benefits to California :

Based on new data regarding the characteristics of California building stock and statewide energy use, the projected benefits of this project are updated as follows:

Updated Baseline

The original baseline conditions and projected outcome were based on national estimates because there was no reliable data for California as a whole. During the course of the Program, the Commission has established a data set for California from which the original estimates can be improved. See Appendix I.

The diagnostics developed in Project 2.3 and the prior work at NIST are intended for air handling units and variable air volume units in larger commercial buildings with central plant systems. The estimated energy consumption for these components is 9328 GWh/yr for Year 2000. The electrical energy used by central plant chillers to provide cooling water to these devices is estimated to be 4872 GWh/yr. The total annual electric use is 14,759 GWh/yr.

Updated Outcome

Colleges, hospitals, and large office buildings are likely to have a large percentage of built-up air handlers and VAV boxes. The total floor area of these facility types was 1573 million SF. Assume that AHUs and VAV boxes that are controlled by controllers capable of having embedded diagnostics serve 50% of the total floor area. Further, assume that if faults are not corrected, they account for 1 kWh per square foot of floor area of energy use by these facilities. The total average annual consumption for heating, cooling, and ventilation for these facilities was about 7.9 kWh/SF in Year 2000. Thus, faults account for about 13% of electric energy used by these end uses for these types of facilities. If Project 2.3 diagnostics were used for 50% of the floor area for colleges, hospitals, and operators were 100% effective in timely correction of faults, the total energy savings would be 786 GWh/yr. Assuming the operators are 50% effective in responding and correcting faults, the savings would be 393 GWh/yr.

Back to Previous Page