Taking control of BAS
This month's panel discusses building automation systems (BAS) and how engineers can meet expectations for these systems, in areas such as installations and retrofits. CSE: How can BAS designers meet an owner's requirements on time and on budget? HYDEMAN: A successful controls project rests on the following pillars: a clearly defined scope of work; a competent and reputable controls contractor;...
This month's panel discusses building automation systems (BAS) and how engineers can meet expectations for these systems, in areas such as installations and retrofits.
CSE: How can BAS designers meet an owner's requirements on time and on budget?
HYDEMAN : A successful controls project rests on the following pillars: a clearly defined scope of work; a competent and reputable controls contractor; and a thorough process for testing and verifying the installed system.
In my experience, the quality of the installing controls contractor is more important than the system that is installed. Specify minimum qualifications for the contractor and the individuals involved in the project, and make sure that these qualifications are submitted for approval.
Critical steps for control system testing and verification include: detailed review of the submittals; spot checks on sensor calibration and loop tuning; functional testing of the installed system and trend reviews of all system points during a two to three week period.
SANTOS : Focus on the fact that this is the brain of the HVAC system, the part of the system that energizes everything, positions all actuators, provides safeties and interlocks and should be thoroughly designed before construction. Do not “punt” this responsibility to the contractor. Research the differences between various types of controllers offered by the manufacturers and respective strengths and weakness relative to network robustness. Investigate the variations between different types of sensors and their correct applications. Write specifications with clear direction on these issues relative to the application where they are being applied.
WICHENKO : Many problems result from viewing controls as a commodity business. One is not simply buying hardware like pumps or drywall screws. Controls contractors provide hardware and software that become integral to the operation of the owner's systems.
By treating controls vendors as equals in a construction contract, the specifier is treating controls as a commodity business based on low price, not best value.
I prefer to specify and procure controls on best value, not low price. Our firm works with the owner to determine the desired features and then we ask the vendors to show how they can provide these features. We have a scoring system to evaluate products. The owner now has a basis for comparison among the prospective vendors and knows in advance what features each vendor can provide. We award controls jobs using a two envelope RFP model that considers features and price. We have found that this model will meet the owner's requirements and will still be on time and on budget.
The other piece of advice I would give is to read “The Contractor's Guide to Change Orders” by Andrew Civitello. I found this an extremely useful book that has helped me write specs to avoid change orders and still have jobs come in on time and on budget.
CSE: Where do you draw the line between consulting engineer and controls contractor/integrator on a project?
WICHENKO : I assume this is a job where the engineer is responsible for specifying the job and the contractor is responsible for installing the job. The engineer is responsible for the high-level architecture, the sequence of operations, the types of panels to be installed and the front-end or Web server to be provided.
I have seen some specs where the engineer has listed how many inputs and outputs are on the panel. This is unnecessary and may restrict bidding. I specify that control for a piece of equipment—say an air handler—must not be split between panels so there is standalone control if there is a network communications fault. The contractor must supply and install a panel of sufficient capacity so this does not happen.
Equipment such as boilers, chillers, speed drives, electric meters and humidifiers now come with BACnet on board. The engineer must now specify these features and then direct the controls contractor to provide a network connection to this equipment. Lighting, fire and security systems also are now BACnet-based. The engineer must specify this equipment and the responsibilities for each vendor to verify that the work was done properly.
This is a necessary but complex task. Our firm has migrated our controls specs to the new CSI MasterFormat Division 25 Integrated Automation specification to simplify this process. Division 25 allows these functions to be put under one division. The job can now be bid as one contract by a vendor who offers all these requirements. It may also be bid as a series of contracts where there are separate direct digital controls (DDC), lighting, fire and security contractor. The engineer needs to specify the integration plan and the responsibilities for implementing that plan. The Division 25 model simplifies this process so trade conflicts and coordination items do not become change orders.
HYDEMAN : All of our controls projects are design-build. In the design-build model, the engineer specifies, reviews and tests the system and component performance, and the contractor provides the detailed design and installation. We don't install control systems and contractors do not choose the quality of the sensors, the sequences of operation and other critical issues of system performance.
We rarely do control drawings. Each manufacturer has different architecture so it would be hard to specify and enforce detailed control system architecture without locking one manufacturer in. We are concerned about performance not the details of how it is achieved. For example, we have a test for the network throughput capabilities that we specify rather than prescriptively mandating the maximum number of low level controllers that can be hung on an MSTP network.
SANTOS: Personally, I draw this line very rigidly. The engineers should engineer and the contractor should install. In reality, this line becomes blurred. Engineers typically have not kept up with the world of BAS and DDC and the rapid change that has occurred here. This particular segment of a typical HVAC system has changed more dramatically than any other portion over the past 20 years and will continue to do so. Current practice has design engineers performance-specifying much of the DDC system and significant portions of the sequence of operation. Practically, this forces most projects into a design-build mode by the controls contractor.
CSE: What's your least favorite memory on a controls project?
SANTOS : Unfortunately, it's hard to choose just one. But in general, I can classify the least favorite memories into a group. These are projects where one single party in the process doesn't understand the issues, execute their role or stay focused on the goals of the project. The result is OK, but falls short of potential due to a single failure by one party. Good BAS/DDC projects don't just happen naturally. They take a committed and educated owner, engineer, contractor and commissioning effort.
HYDEMAN : My least favorite memory occurred when I was working so hard to get a contractor to perform that I not only blew my fee but was donating my time to fix the system. After twice rejecting the control programming submittals, which had no relation to our sequences, I called the branch manager and had them pull the programmer. This was just the beginning of a cascading sequence of events that evoke Bill Murray's trials in the movie “Groundhog Day.”
WICHENKO : My least favorite memory is being embarrassed because the features that I thought I was getting were not provided. I blame myself for not having the knowledge about the systems and the specifying tools to avoid this from happening. I have spent a lot of personal time to educate myself so this never happens again.
CSE: What's your most rewarding experience on a controls project?
WICHENKO : Our firm recently completed a multi-million dollar PLC/DDC retrofit job in a remote location. We had one minor change order. The job went in very smoothly. Both the owner and the contractor were happy with the results. The job was extremely difficult to specify and bid out, but it worked out well.
SANTOS : My most rewarding experience involves taking an organization or institution and changing the way they procure, design, accept and maintain their DDC systems. This typically has involved working with operations personnel, engineers, project managers and contracting personnel to produce a set standards (specifications, drawings, sequences) for common HVAC systems and the subsequent education of users of the documents (engineers, commissioning firms) for various projects. The development, use, education and enforcement of this BAS/DDC master plan takes time and resolve to implement successfully, but over time can change the current approach: business as usual, hope and pray things come out right.
CSE: Where would you advise manufacturers to invest R&D dollars for BAS hardware or software?
HYDEMAN : I would love to have self-tuning loops that really work, adaptive control algorithms, which have been in use for decades in industrial systems, and automated fault detection and diagnostics. However, before I spent money on any of these things, I would encourage the industry to develop and adopt a rigorous certification program for control technicians. The industry has shot itself in the foot by failing to provide qualified and well-trained technicians. They are also under bidding jobs, which forces them to cut corners. In addition to training the technicians, the project managers should be sent to business school.
SANTOS : Making products easier to use, install and setup. Invest fewer dollars in the so-called open protocol hardware/software area and put more emphasis on robust controllers that are easy for their installers to configure and complete. Focus on developing more effective training and documentation for installers and users. If you focus on system openness, develop and install truly open systems. Open protocol is a necessary precursor to open systems, but far too often, promises are made for open systems while only delivering a limited open system uses an open protocol. One needs to compare the functionality of a single manufacturer solution to a truly open system. A system is not open where functionality is limited across platforms and proprietary software or hardware keys are required to operate and configure these systems.
WICHENKO : I believe that BACnet is the protocol of choice for controls equipment. USGBC LEED Silver certification is now the de-facto standard for new construction. Owners will want integrated, interoperable systems to do part-load occupancy control, demand-response strategies, measurement and verification and continuous commissioning to get LEED credits. BACnet is a continuously evolving standard and manufacturers must keep up with these protocol changes to meet changing owner requirements.
Some vendors have taken the approach that control devices can be relatively simple with the integration being done at the server level. It is my view that owners want Web-enabled, networkable equipment that will run for 15 to 20 years in a standalone mode when there are network communications faults and server crashes. I believe that distributed control is the better model for our industry and is more consistent with the need to have secure, fault-tolerant systems in this post-9/11 environment.
CSE: What do you deem as low-hanging fruit for most energy-efficiency retrofits that involve BAS?
HYDEMAN : Zone-based reset for temperature and pressure.
WICHENKO : Re-commissioning DDC systems is a cost-effective way to gain savings. Many utilities offer grant programs for this work.
The other important item is utility meter reading through the DDC system. Converting gas, water, electric meters to DDC is a relatively simple job now. Even early legacy systems can accept a pulsed input from a meter that can be accumulated in a panel. This information gives valuable information to maintenance on utility consumption on a daily basis. If there are changes from day to day, this will alert maintenance to go check for leaks, faulty equipment etc. This data can be passed up to the enterprise level using Web services for further analysis or for making real time pricing buys.
Mark Hydeman , PE, FASHRAE
Principal, Taylor Engineering, Alameda, Calif.
Jay Santos , PE
Principal, Facility Dynamics Engineering, Columbia, Md.
Grant Wichenko , PEng
President, Appin Assocs.Winnipeg, Manitoba