Green design vs. green performance

The latest version of LEED 2009 (v3) requires high-performing buildings to perform up to even higher standards, but key issues remain unaddressed.

By Patrick Lynch, Associate Editor, and Michael Ivanovich, Editor-in-Chief October 1, 2009

CSE: Does LEED v3 do a sufficient job of certifying a building’s level of sustainability? Why or why not?

Jay Enck : First let’s define sustainable. Sustainable means zero impact on the environment where energy used does not exceed energy generated, water consumption does not reduce availability, materials have zero-embodied energy and do not deplete natural resources, and facilities’ indoor environmental quality promotes positive physiological and psychological responses. Based on the above definition, the current version of LEED does not result in sustainable facilities. It does provide a starting point toward achieving sustainability and provides the basic principles owners and their operation, maintenance, and project teams can follow on a path that will reduce environmental impact. The industry first has to learn the basic principles to advance to the end goal.

Don Millstein : LEED v3 is a large step in the right direction for sustainability. The increased emphasis on measurement and verification (M&V), along with a rationalization of energy conservation measures and cost parameters, will go far toward facilitating new levels of analysis and design for sustainability.

Hernando Miranda : LEED predicts sustainability in a project, but only based on the measure it includes in its rating system. By itself, it is not sufficient. There are many sustainability features which LEED has yet to address; measures that have been around for a long time. Building size is one measure. If owners can be convinced that a manager’s office can be 100 sq ft, instead of a more typical 120 to 150 sq ft, and open office plans were used for office staff, it could be possible to reduce a building volume by 10% to 20%, or more. The size of building systems and envelope would be reduced, not only providing significant cost savings, but environmental ones as well.

Reducing the amount of finish materials is another measure that is not included in the LEED rating systems. As an example, it is possible to achieve a Platinum building with one using a significantly greater amount of finish materials (e.g., custom wood walls, multiple carpet types), while the other consists of concrete masonry unit walls and concrete floors. The finish-intensive building can include cooling system (even if it is not required), while the other does not (even if it is required to include cooling energy use). The above example is a real-world case, and a comparison of a few differences between two highly regarded Platinum projects.

CSE: Why is there such a fundamental disconnect between design and construction of new buildings, energy systems, and envelope, and how they actually perform?

Mark Frankel : By far the most common use of energy modeling, including [ U.S. Green Building Council ] LEED modeling, is to compare a proposed design to a theoretical code building. The inherent goal of this modeling is to demonstrate that the proposed building will use less energy than the theoretical code building (or to compare alternate performance strategies), not to accurately determine actual energy outcome. Neither LEED nor energy codes regulates actual energy performance outcome; so there has been no incentive for modeling accuracy with respect to actual performance outcome. The one exception: The Living Building Challenge, by CascadiaGBC , requires one year of data to prove net zero energy performance.

Enck : Often it is a lack of quality control throughout the project delivery process combined with inadequately trained facility operation, and a lack of monitoring performance. In our existing building and forensic work, we often see:

  • Disconnects between the design disciplines that compromise performance

  • Design and construction coordination issues that compromise maintenance accessibility necessary to maintain building performance

  • Vague construction documents that allow construction practices that are cheaper to install but more expensive to operate

  • Operational problems due to insufficiently trained operation staff

  • Zero monitoring of performance by the owner.

Millstein : It’s difficult to design an energy-efficient building to perform in a precisely anticipated manner, due to all the estimations and assumptions that go into the mix. Since the number of occupants, job functions, space use, typical lighting/HVAC requirements, plug load, and other factors can vary greatly, designers will typically use an industry average that does not take into consideration specific end-user requirements or significant deviations regarding traditional use of the facility. Also, manufacturers’ specifications are at times overly aggressive, or based on an entirely different set of assumptions that further widens the gulf between estimated and actual performance.

Miranda : Actual construction does not necessarily reflect what can be drawn on a computer, or analyzed by software programs. Weather is another factor; actual conditions vary from various weather assumptions made by software. Weather assumptions bound large geographical areas. Buildings with very different microclimates can be forced to use the same weather assumptions based on energy modeling requirements.

CSE: How can energy modeling programs improve how they accurately predict a building’s actual energy usage?

Enck : Having more robust energy modeling programs that include better daylighting modeling features, HVAC system configuration capabilities, and control interaction would be beneficial. The main issues with energy model accuracy are the energy modeler’s input assumptions and the limits in being able to predict how a facility will actually be used and operated. Owners and users typically do not fully know or communicate all of the connected loads in a facility. Accurate energy modeling requires measurement of actual conditions, which is only completely accurate at the instant the data were measured.

Millstein : Normalizing the data allows benchmarking against other users to be more effective. However, if the usage within a specific facility deviates substantially from the standard for a known reason, the numbers should be adjusted for this deviation for a proper comparison. If one assumes that a specific facility has computer rooms that are double the traditional required square footage, the data will need to be adjusted to reflect the increased energy required for computing power as well as HVAC to properly benchmark the facility.

Frankel : Actually, the modeling programs are sophisticated enough to accurately predict performance outcome, given the right inputs. Although there are some anomalies in each modeling program, the bulk of the ”prediction error” has to do with the fact that we really don’t know how the building will be operated or occupied at the time of design, or whether the systems will be installed or operated correctly. In general, about half of the values on which the energy modeling calculations are based are simply assumptions about how the building will be used or operated, unrelated to physical or performance characteristics of the systems themselves. Until we have meaningful feedback loops to provide realistic and accurate assumptions about schedule, system operation, occupancy patterns and use habits, maintenance, and plug loads, modeling accuracy will continue to suffer.

Miranda : They can be improved by focusing on comparisons of differential performance between different MEP systems and envelope designs, rather than as absolute predictors of energy use. Comparative results are the best method for deciding whether a potential system or building improvement should be included in a design. In terms of actual energy usage, modeling programs should include a built-in method for entering actual energy use, to allow comparison to predicted results. For example, comparing the energy usage savings of specifying higher efficiency pumps for an HVAC system, might not be as cost effective as specifying lower solar gain glazing. The glazing might have a higher first cost, but based on a cost-per-energy unit saved, the glazing might be a better choice than the pumps or vice-versa.

CSE: How can LEED v3’s new M&V requirements be met? Which loads are most important to monitor separately to inform energy modeling improvements?

Frankel : The new LEED M&V protocol is a work in progress, and it remains to be seen what the program will focus on. In the past, LEED has focused on a set of very detailed data on building performance that cannot be reviewed by a layman. Since there are no simple cues to the building owner/operator about whether the building is performing well or poorly, there is no apparent incentive to hire an engineer to really analyze the detailed data being collected. What is needed is a set of simple performance indicators that provide the cue when perhaps a deeper look at the data is warranted. In the absence of this “cue,” LEED M&V is an expensive and complicated metering protocol generating mountains of data no one ever looks at.

Millstein : Installing all types of submeters—electric, water, gas, etc.—is key to fulfilling M&V requirements. Beyond metering HVAC and lighting, data centers and intensive plug loads are big users of energy and need to be metered not only for M&V, but as part of the overall energy management program. The old adage, “You can’t manage what you don’t measure,” demands a better understanding of where energy is being used both efficiently and inefficiently. This energy visibility also allows facility professionals to identify where budgeted dollars for energy conservation should be spent first and to verify the savings for future investment.

Enck : There are two options using a calibrated simulation (IPMVP Option D) or energy conservation measure isolation (IPMVP) for a period of at least one year post occupancy. Option D can use the energy model used to predict the facilities performance during the design phase and turn it into a calibrated model by updating the assumptions with actual system parameters, occupancy, and weather data. Options B requires monitoring separately lighting, process, plug, and HVAC system energy consumption. If the project can divide these loads in the electrical distribution then the number of metering points and cost of metering can be minimized. Monitoring of the different load types, including plug load, provide valuable insight into optimizing the buildings operation.

Miranda : The LEED developers can work with the M&V experts by providing data for their use, and helping develop detailed methods of improving load monitoring. Plug loads are a growing energy usage in buildings. Now is the time to start requiring these plug loads to be broken down into smaller components. The United States Green Building Council/Green Building Certification Institute should make the credit for an enhanced version of M&V a prerequisite with the next version of LEED.

CSE: People look to LEED for energy performance; meanwhile LEED has adopted ASHRAE Standards 90.1 (energy efficiency), 62.1 (ventilation and IAQ), and 55.1 (comfort). Is part of the performance problem the complexities or interactions among these standards?

Miranda : Yes, the three standards do not necessarily interact with each other well. ASHRAE 90.1 should require compliance with ASHRAE 55.1 comfort charts. ASHRAE 62.1 ventilation results should also be required for compliance with ASHRAE 90.1. ASHRAE 90.1 is not set up to model naturally ventilated buildings well; a building that does not require installation of a cooling system is required to model one and include its energy usage. ASHRAE 62.1 and 55.1 include requirements for natural ventilation. ASHRAE 90.1 is the standard that needs the most work to integrate the three standards effectively.

Millstein : The underlying standards can add some performance problems since ultimately there might be a trade-off—not only based on cost, but performance when designing the building to meet all the standards.

Frankel : There are certain circumstances where these standards set up design conflicts. A design conflict example would be when minimum ventilation rates drive up outside airflow, thereby increasing heating/cooling energy use, or when a single system is used to deliver heating and cooling within a building simultaneously. These conflicts tend to be “self-inflicted” by designers. Well-designed systems using strategies like dedicated ventilation systems with radiant heating and cooling can meet all of the standards without conflict.

Enck : To achieve sustainable, high-performance buildings requires guidelines and standards that provide design and construction criteria. ASHRAE provides codified standards at the request of our government. The raising of the energy efficiency bar, while also maintaining ventilation and comfort criteria conducive to occupant productivity, does increase the stringency of these standards and their complexity. I have not observed any conflicts within these standards which interfere with building performance.

CSE: What’s at the core of the performance issue? Is the real world or construction and operation of buildings too complex to model accurately based on tools and information (data) available today?

Enck : There are multiple reasons including not clearly documenting the end-goal; incorrect energy modeling, uncoordinated design, poor workmanship, improper operation, and maintenance. Energy models typically provide only an estimate of performance. Unless you verify the energy model assumptions against actual conditions, including how leaky the building envelope is and actual weather conditions, the results will never match the actual. On the other hand, a building working perfectly at the beginning of occupancy can have its performance quickly degrade from operator error.

Frankel : The heart of the matter on building performance is that no one pays attention. Once a building is in place, there is seldom any effort to follow up on how the building is performing. Although nearly everyone knows what kind of gas mileage their car gets, hardly anyone knows whether their home or office building is good, bad, or indifferent from an energy performance standpoint. In addition, there are almost no mechanisms in place right now to provide this feedback. Tenants typically have no feedback on how their behavior affects building energy use. Energy modeling is a great tool to predict relative performance under specific conditions. But whether modeling is accurate or not has little relevance on building performance. If we could compare the performance of buildings by two different architects or developers, there would be immediate pressure on the design community to pay attention to performance outcome, and to deliver better buildings.

Millstein : It’s a two-fold issue revolving around variability. The first issue is the variability of estimates based on typical usage or assumptions used by manufacturers when developing their specifications. The second issue is the variability of how the building actually operates versus original estimates. Once a building is occupied it becomes a “living entity” that will change on a daily, if not hourly, basis in terms of energy consumption patterns and efficiency. The interaction of people and a building is extremely difficult to model.

CSE: Are engineers who work on LEED projects facing additional litigation or liability risks if the buildings fail to meet energy projections?

Miranda : Yes, if the USGBC/GBCI assume they can use actual energy usage as a means to decertify a project. Engineers and energy analysts do the best they can with the software they use. The software does the best it can using equations developed by other organizations, such as ASHRAE, and weather data provided by third parties.

Frankel : Modeling accuracy will become an increasingly important issue, but not because the lawyers are circling. As it stands, it would be impossible to prove a building performance case against a designer, unless there were true errors and omissions. If you try to sort out a lack of performance issue in a typical building today, it is impossible to separate the energy impacts of the design from the impacts of construction, operation, maintenance, and tenant behavior. Energy modeling is based on a specific set of assumptions about the materials and equipment being installed and working exactly as designed, for certain specific hours each day, under specific temperature, humidity, wind speed, and sun conditions, with a specific number of occupants behaving a specific way at all times. LEED and energy codes in general require that buildings demonstrate performance above a theoretical baseline, not at a specific energy consumption level.

Enck : When there is a contractual performance requirement for energy/water efficiency or LEED certification not achieving the performance requirement, that will increase the risk of litigation. One of the first examples is a 2007 Maryland circuit court lawsuit: Shaw Development sued Southern Builders for $635,000 in lost state tax credits for failure to achieve USGBC LEED-Silver Certification. Most standard agreements do not address green/energy efficiency and leave open who is responsible for achievement of a green rating, or what happens if it is not achieved. Reportedly there are many such cases in the litigation process, and because it requires a team effort to achieve [certification], the team is at risk.

CSE: What role can consulting engineers play in the long-term performance of the systems they design? Besides engineers, who else should building owners look to for help with long-term performance?

Miranda : Engineers and owners can both be helped by including M&V experts as part of the building design process, and continued performance analysis after a project has been constructed. Engineers and architects should be required to continue their involvement with a project long after construction is complete. This way they can learn what works well and what needs improvement, and improve the design of future buildings.

Enck : Building performance is a team effort in which engineers are an extremely important part. Owners of successful high-performance buildings have one commonality they properly maintain—measure performance and correct deficiencies on continuous bases. To achieve this they have either maintenance-centered, engineering-educated employees and/or commissioning authorities that measure, evaluate, document, and correct deficiencies. For new facilities, starting the commissioning process ahead of hiring the design team and following it through the life of the facility can help ensure performance is achieved for the life of the facility.

Millstein : In their vital project development role, consulting engineers need to take a longer-term view of building operations, while at the same time balancing the financial aspects. Developing a long-term vision for building performance should involve a committee that includes not only the architect and engineers, but facility professionals, building occupants who will be responsible for much of the energy load, and representatives from various vendors supplying equipment to the project. Factoring all these inputs into the equation will not only improve the longer-term design view of the building’s usage, but provide new ideas that could dramatically impact its performance.


H. Jay Enck, LEED AP

CEO Commissioning & Green Building Solutions Inc., Buford, Ga.

Mark Frankel, LEED AP

Technical Director, New Buildings Institute, Seattle

Don Millstein

President/CEO, E-MON LLC, Langhorne, Pa.

Hernando Miranda LEED AP

Principal and Sole Proprietor, Soltierra LLC, Carlsbad, Calif.