HVAC Controls
By Philip Haves
Applications
Conventional HVAC controls consist of cascaded single-input controllers, generally proportional-plus-integral (PI) feedback controllers at the lower levels ('local loop control') and controllers based on heuristic logic at the higher levels ('supervisory control'). Current systems do not include the more sophisticated types of control found in other industries, e.g., multiple input, multiple output controls, whose skill requirements exceed those found in the buildings industry. In general, packaged roof-top air-conditioning units have their own, dedicated controls whereas the 'built-up' systems typically found in larger buildings have programmable controls, usually configured using a graphical programming language.
Best practices in HVAC controls include:
Sequences of operation produced by the mechanical designer in the case of one-off system designs or selected from a set of standard, previously tested sequences for routine design
Commissioning by an independent third party provider hired at "arms length" by the architect or the client, rather than the mechanical contractor
Performance monitoring using either the building control system or a separate system for data acquisition, archiving, visualization, benchmarking, performance analysis and diagnostics,
Full documentation for all systems and comprehensive training for operators and other facilities staff, with unrestricted access and all programming tools required to make changes throughout the life-time of the system
Anecdotal evidence indicates that it has become common for mechanical designers not to provide sequences of operation and to rely on the controls contractor, in practice a technician, to generate the control program. The programming used in a new project is typically based on previous projects involving the same types of system, a practice that works to the extent that the current project and associated occupancy and user needs are similar to the previous project and the extent to which the original sequence of operations was adequate. Further anecdotal evidence indicates that this practice has become engrained to the point where, in cases where sequences of operation are provided by the mechanical designer, these sequences are usually ignored in favor of previously-used sequences.
Specific sequences of operation are likely to be provided when the mechanical design is innovative in some way; ignoring these sequences can lead to a combination of frustration, delay, and poor system performance (with adverse consequences for occupants). If the project is commissioned, it typically falls to the commissioning provider to resolve inconsistencies between the design intent and the sequences of operation as implemented by the contractor. There are exceptions to this pattern but it is pervasive enough that the owner or the owner's representative needs to take steps to ensure that it does not occur if good performance is desired. Keys to ensuring a good outcome are explicit specifications and careful selection of the controls contractor and the commissioning provider.
To address the common situation of conventional systems, ASHRAE commissioned a research project (1455-RP) to develop standard sequences of operation for air-based secondary HVAC systems (Hydeman et al. 2015). The maintenance and dissemination of these sequences, and any sequences developed in the future, will be under the auspices of the new ASHRAE Guideline Project Committee (GPC) 36 High Performance Sequences of Operation for HVAC Systems. It is expected that the major controls vendors will create libraries of these sequences of operation, anticipating that designers will start to specify them in controls bid packages. In addition to these sequences being directly usable in many projects, they will provide good starting points for sequences for somewhat less conventional systems. The deliverables from1455-RP are available on the GPC 36 website (ASHRAE GPC n/d). It is expected that validation and demonstration of the 1455-RP sequences and development of standard sequences for primary HVAC systems will be addressed in future research projects. It is anticipated that sequences for innovative systems will be developed as and when these systems mature and become standard practice.
Despite the success of delivering automated facades and dimmable lighting in a large class A office building over a decade ago, e.g. the New York Times headquarters building (Lee et al. 2013), and the positive results from other underlying field studies that support these potentials, these controllable facade and lighting systems are still only occasionally implemented in buildings today due to costs and the complexity of integrating sensors, controls and dynamic system elements, which are compounded when HVAC controls are also integrated. By default, facade and lighting systems use separate controls, so integrated control requires data communication among the controls. Possible technical solutions are (i) to use an integration platform, such as Tridium's Niagra framework, (ii) to use BACnet to implement simple data interoperability or (iii) to use a single controls platform that has the versatility to meet the control requirements for the different systems. Separate energy information systems (EIS), which may link data from control systems and energy metering systems, and provide archiving, visualization and analysis capabilities add further complexity and increase the attractiveness of integration platforms, although some HVAC controls vendors provide EIS functionality in add-on software products.
Commissioning can substantially improve the performance of poorly installed control systems. It typically consists of two stages: pre-commissioning and functional testing. Pre-commissioning involves checking the wiring from controllers to sensors and actuators and verifying that the behavior of the sensors and actuators is as specified. Functional testing involves changing set-points or overriding control signals to actuators and verifying that the response of the system or components is as expected. It also involves testing start-up and shut-down sequences and verifying the system behaves safely in the event of malfunctions. Commissioning is valuable in that it it uses systematic testing to reveal faults and other operation problems at or before the start of occupancy; if commissioning is not performed, faults appear in an ad hoc fashion over time, when they are more difficult to diagnose or remedy.
Building control systems typically only include the sensors required by the sequence of operations. Monitoring of performance - energy or indoor environmental quality - requires additional instrumentation, archiving and visualization to analyze performance and detect and diagnose problems, even manually. Installation of electricity sub-meters on key circuits or enabling power read-outs from variable frequency drives and complex equipment, such as chillers, provides valuable information that allows performance to be assessed. ASHRAE Guideline 13 Specifying Building Automation Systems includes an addendum on instrumentation for performance monitoring and ASHRAE Guideline 22 Instrumentation for Monitoring Central Chilled Water Plant Efficiency also includes analysis procedures for characterizing system performance.
More advanced control technologies are starting to be developed in response to the emerging use of thermal storage systems and weather forecasting. Thermal storage systems include tanks of hot water, chilled water or ice and thermally active building systems (TABS), such as radiant slabs. The desire to optimize the operation of such systems has created a need for predictive control; one option is model predictive control (MPC) (Ma et al 2012). MPC uses a simplified model of a building or a central plant, together with weather and utility price forecasts, where available, to determine an optimum control strategy for a period of a few days. The strategy is continuously updated in response to the actual performance of the building or plant and updates to the forecasts. This process minimizes energy costs by reducing under- or over-charging of thermal storage. The motivation to adopt MPC will increase as buildings are incentivized to adapt to increasing volatility on the electric grid caused by an increasing fraction of generation from renewable sources. However, the limiting factor is likely to be the skill level of designers and facilities staff. Reliable performance can only be obtained if the operating staff can attain, and maintain, sufficient understanding of the systems and their controls.
Campuses, particularly those belonging to organizations with technically-based management and decision-making, offer the most favorable circumstances for the successful adoption of new technologies, including advanced control systems. The scale of campuses justifies the employment and training of the specialists required to enable the evaluation and adoption of new technologies.
One obstacle to the implementation of high-quality sequences of operation for one-off buildings is a lack of controls design analysis tools adapted to HVAC, particularly for use in schematic design and in early design development, before the controls vendor has been selected. Generic tools such as MATLAB (MATLAB n/d) have not found favor in the HVAC industry and are rarely used. The controls modeling capabilities of whole building energy simulation tools such as eQUEST (eQUEST n/d) and EnergyPlus (EnergyPlus n/d) are largely limited to simple supervisory strategies and do not address local loop control. Similarly lacking are the capabilities to address integrated control of HVAC, facades and lighting. One approach is to use more modular simulation tools, such as TRNSYS (TRNSYS n/d) or emerging tools based on the Modelica modeling language (e.g. Wetter et al. 2014), though, as with MATLAB, these tools do not scale easily to the whole building level and require skills not often found in the buildings industry.
In addition to the lack of a control strategy design tool for buildings, the current process for implementing control strategies is largely manual and suffers from a number of problems that result in buildings failing to perform to their technical potential unless remedial measures are implemented as part of a commissioning process. The multiple disconnects in the current process can be overcome by the automated tool chain illustrated in Figure 1, which is still in the concept stage for buildings but is based on methods that are already firmly established in the semiconductor and automotive industries (Sangiovanni-Vincentelli and Ricorsi 2010; Di Natale and Sangiovanni-Vincentelli 2010). The key is to represent the sequence of operations produced using the design tool in a machine-manipulable form, with the format defined by an open standard. Hardware selection, network design, and generation of functional tests are performed automatically and the distributed control program is then generated, uploaded and tested automatically. The main benefit, particularly for more innovative designs, is that it is possible to proactively guarantee that the sequence of operations produced by the designers is correctly implemented in the actual control system, which is often not the case in the current process. Once developed, this process could be implemented both by the established controls vendors and by new entrants to the controls market.
Figure 1. Future tool chain for automated design and implementation of building controls
A number of research organizations are investigating new, open architectures to support a wide range of operational functions in buildings and beyond, including integrated control, automated diagnostics and grid transactions. If, and when, any of these efforts come to fruition, they will enable a number of the limitations discussed above to be overcome much more easily and efficiently, potentially transforming not just the operation of individual buildings but the role of buildings in relation to the enterprise and the grid.
Economics
Control systems are enabling technologies in the sense that the cost of the mechanical equipment and active facade components that they control typically exceeds the cost of the control system and so it is difficult to consider the cost of the control system in isolation. Installation costs generally exceed the cost of the control hardware and a major component of the installation costs is the wiring of sensors and actuators to the local controllers. Robust wireless communications could reduce these costs significantly but existing wireless technologies have failed to displace hard-wiring. Sensors and actuators with digital interfaces connected by wired networks could also potentially reduce installation costs but have also failed to have a substantial impact on the HVAC market, in part because there are multiple standards. Proactively commissioning control systems is significantly more cost-effective (and less disruptive to occupants) than piecemeal reactive troubleshooting and repair.
Institutional requirements & capacity
Cyber-security has become an important issue for building control systems, particularly since a recent security lapse at Target Corporation was attributed by several commentators to the lack of segregation of the financial systems from the HVAC controls network, though not to the HVAC control system per se.
As noted above, the relative lack of special control engineering and related skills among designers, contractors and operators is expected to limit the adoption of more advanced control techniques and systems. Organizations that have invested in the recruitment, training, and retention of staff will be better placed to exploit emerging advances in controls technology. Commissioning providers can be looked to as one source of training. More generally, close attention to specifications is particularly important for control systems and use of tools such as ASHRAE Guideline 13 Specifying Building Automation Systems can be part of to a quality assurance process that can help prevent the control system being the Achilles heel of commercial building performance.
References
Di Natale, M. and A. Sangiovanni-Vincentelli. 2010. "Moving from federated to integrated architectures in automotive: the role of standards, methods and tools." Proc. IEEE 98(4).
ASHRAE GPC 36. http://gpc36.savemyenergy.com/public-files/
EnergyPlus. http://www.energyplus.gov
eQUEST. http://www.doe2.com/equest/
Hydeman, M., S. Taylor and B .Eubanks. 2015. "RP-1455 and Guideline 36 control sequences and controller programming." ASHRAE Journal.
Lee, E.S., L. Luis, L. Fernandes, B. Coffey, A. McNeil, R. Clear, T. Webster, F. Bauman, D. Dickerhoff, D. Heinzerling, and T. Hoyt. 2013. "A post-occupancy monitored evaluation of the dimmable lighting, automated shading, and underfloor air distribution system in The New York Times Building." LBNL Technical report, 2013. LBNL-6023E.
Ma, Y., F. Borrelli, B. Hencey, B. Coffey, S. Bengea, and P. Haves. 2012. "Model predictive control for the operation of building cooling systems." IEEE Transactions on Control Systems Technology, 20 (3), 796-803. http://escholarship.org/uc/item/2hz3g57z
MATLAB. http://www.mathworks.com/products/matlab/
Sangiovanni-Vincentelli, A. and C. e Ricorsi. 2010. "The EDA Story." IEEE Solid-State Circuits Magazine, pp. 6-25, Summer.
Wetter, M., W. Zuo, T. S. Nouidui and X. Pang. Modelica Buildings library. Journal of Building Performance Simulation 7(4):253-270, 2014
TRNSYS. www.trnsys.com