Meeting Growing Need for In-House Technical Expertise

Hi Patrick: In addition to myself, we have a High Performance Coordinator who performs most of our technical work: daylighting studies, LCA, and assisting teams with rating system credits and documentation. We also do some lightweight energy studies using Cove.Tool, which I am lukewarm about but that is another topic by itself. I originally intended to train more staff to perform these studies within teams, but that has not proven feasible. It is faster and more accurate to have the HPC do the studies. She is familiar with the software, knows the tricks and workarounds, and can turn around results to the teams in a short time. Kristian

Hi Patrick, We typically work on larger or unusual/complex project types, so this may not apply to you. In addition to me (flit between projects to offer workshops, strategy, and some technical advice as well as directing research, doing some marketing, and some other stuff) we have a full-time building performance specialist (Chris Savage) who trains others and performs most of our daylight and lightweight energy analysis, similar to the HPC that Kristian mentioned. We have also found it more likely that simulation will get done by the specialist, though we train team members to do it in case they can, and also so they know what can be done and become conversant about it. All of the early whole-building energy modeling is done by third party specialists because we work on larger or complex projects. We have another person (Justin Schwartzhoff) who leads our embodied carbon efforts, though only with a small part of his time. We do embodied carbon analysis internally on all projects. We also have three Applied Research Teams that gather interested LMNers to learn and help projects with some other technical topics. Chris and Justin each lead one of the teams. One of my roles is developing in-house talent and expertise, so I meet with several individuals regularly to help them grow their sustainability knowledge. -Kjell

Hi Patrick, This is a question we have always had some trouble with. There are certainly some more advanced sustainability analyses that take an expert to perform, but our office policy has long been to train everyone in the use of basic tools so they can do it themselves on each project. We've been heavily pushing adoption of cove.tool and asking all teams to do simple energy models in it as their design progresses. Unfortunately, results are mixed because not all projects have a strong sustainability component and this can fall by the wayside. As the Sustainability Coordinator, I still end up doing or at least helping out with a good number of the models. And if we need to use other software like Ladybug, that also lands on my plate. So I don't know what the best solution is. I think for us the model of a hand-holding Sustainability Coordinator helping a staff with basic competence is sort of working. But we do outsource more advanced whole-building energy modeling to our consultants when things get beyond cove.tool's capabilities. Kristian, I'm curious to hear your lukewarm thoughts on cove tool. It has a lot of things I like and a lot of things I am frustrated with, but so far it is the only platform I've found to be user-friendly enough for the whole office to use, which for me makes it worth the hefty price tag. Hope to keep this conversation going--it is such an important topic and one I think everyone is figuring out as they go...
  Misha

This is a little sidebar but whether you use inhouse expertise or external consultants, ASHRAE Standard 209 has good guidance on several things including modeler credentials, climate and site analysis, an energy charrette, energy performance baseline and goals, and quality assurance. We have an ongoing monthly 'sustainability roundtable' organized by the education subcommittee of our sustainability forum. In 2021, we've had two sessions so far on building performance covering a wide range of topics, including multiple simulation examples. Ramana Koti Senior Associate, BEMP, LEED Fellow Responsive Design LordAeckSargent.com

Here's my long rant on Cove Tool. We have used both Sefaira and Cove Tool. Our experience with Sefaira is now about 3 years out of date, may not be relevant any more with recent changes. For Cove Tool, I am finding its EUI predictions to be "directionally accurate"; within 1 or 2 EUI of more detailed models. What I am not having success with is: 1. The daylighting results. It's very pessimistic on north-facing windows and the results do not match our DIVA/Climate Studio results. Using it, I cannot get a 900sf classroom to 55% sDA even if I make the entire north wall a big pane of glass. It says I can never achieve more than 1 LEED daylight points for a school building in the PNW. So I don't use the daylighting much and we use Climate Studio instead. 2. The optimization tool. They recently changed the way this works and it's much less useful to me now. It's harder to set up custom energy options to run. And, it appears our WA and Seattle energy codes have hit the point of diminishing returns; CoveTool has a hard time suggesting energy improvements because the code compliant options are already so efficient. Mostly what I have been able to use it for is as a quick shoebox model at the start of design, and maybe run some different building massings through it for comparison. Even then, I'm finding the massings vary by about 1 or 2 EUI, not significant enough to justify the selection of one option over another. Even though the EUI results have aligned with the more complete energy models later in the projects, the load distribution looks a lot different than what our engineers model, which makes me less confident in it. We had similar issues with Sefaira in the past. I think one of the biggest problems is that these programs are built around office building occupancies, and they aren't really set up to deal with the variety of space types and system types in K12 and thus I'm left feeling it is oversimplified. It's probably more useful if your teams are doing a variety of project types, so today you're looking at a medical tower, tomorrow a convention center, university classroom building the day after. Because our firm does primarily K12, we pretty much get the same results from the shoebox model every time. It's never provided that "Aha!" moment where a magic combo of options drops the EUI by a significant amount, like it promises. TLDR CoveTool creates some nice imagery and gives our early designs some legitimacy, but it's not really informing our decision-making in a meaningful way. If others are having more success using it, I'd love to hear from you.

This topic comes around again and again!   We're big believers in the use of simulation to inform design choices. In our 2020 reporting for the 2030 Commitment, we had energy models for 98% of our floor area and embodied carbon models for 51% of our 2020 design portfolio floor area--this was a high-water mark for us. But we are increasingly aware that usually we already knew what the simulation was going to tell us--use glass, but not too much; enough with the gratuitous aluminum, already...  So simulation tools are great, but we can most of the way there by not doing stupid stuff.  This comes from an inherent trade-off:
With any sophisticated tool (energy, carbon, or daylight model) the challenge is that it takes a lot of time-on-task to understand the limitations, the tricks, and the workarounds. When we try to devolve training in these tools to general architectural staff, we find that the 'easy' tools tend to convey a false confidence in the results (and different users can get different results for the same building entered different ways), and the 'hard' tools tend to require someone to become a specialist and let go of being a general member of the architectural staff.

We have sometimes been blessed with having a particular staff member who brings deep knowlege and expertise about one tool acquired at a previous job (say, having an architecture degree but having worked full-time on energy modeling for a consulting firm or at a research position) but wants to be embedded in a design firm.  That's pretty rare but wonderful when it happens.

Often, the best we can do is to cultivate one staff member for each tool (say, Tally or Lightstanza) who then becomes a 'coach' to design staff members who learn that tool and apply it on their own project, with the "ninja" checking their work.

In general, though, I find it more useful to just tell designers to start with best practices, as listed below, and if they want to deviate, place the burden on them to do the calcs of have them done.

• Energy: Take the prescriptive requirements of ASHRAE 90.1-2016 as the starting point for envelope, glazing, etc. even if the jurisdiction you're designing for has less stringent requirements.  (If you are working in a place with more stringent requirements, congratulations.)
• Daylight: Start with a Window-to-Wall Ratio of 30% and make no place you'd ask someone to sit for more than an hour be further from the glazing than 2.5H away, where H is the top of glazing height.  If your designer want more glass, then they (not you) will have to do the hard work of proving that it actually provides measurable benefit without imposing unreasonable financial and energy cost.
• Carbon: "Use materials. Not too much.  Mostly wood."  Tell your structural engineer at the start they have to use low-carbon concrete and they have to verify that their mix will be available through the likely redimix suppliers. Scrub your firm's palette of allowed insulation materials of high-GWP options--just make selecting the bad stuff not an option.
• Water: At project kickoff, see if you can give the facility manager a 0.8gpf toilet (like, say, Niagara Stealth) for their home.  $100. Pooping is believing.  For commercial buildings with cooling towers, focus on reducing cooling use as this tends to completely dwarf the water consumed by 'domestic' uses.  Tell your landscape architect at project kickoff that you're deeply sorry but there's just no budget for an irrigation system, not even one that uses captured rainwater, so they'll have to live with plants that won't need it after establishment.
• Air.  At project kickoff, tell your HVAC engineer that they've got to provide 40 cfm per actual person expected in the building (not code-rated occupancy limits) and provide MERV 13 filtration. After they stop sputtering about how this will destroy energy performance you can tell them there are plenty of projects exceeding 70% savings in 2030 that do this. Make the use of ventilation air energy recovery systems the default for your firm's projects.  If you have any high-ventilation zones (e.g. labs) tell the HVAC engineer that they have to implement an air cascade so they put deliver oodles of fresh air to places like offices and meeting rooms and then use this as the make-up air for the labs.  And if you do labs, challenge the Owner to walk their existing buildings and see how many fume hoods are actively being used for their intended purpose, and challenge how many they really need; see if ductless fume hoods are an option and put your client in touch with those that are already using them.
• Airtightness: A building with really good air sealing and so-so equipment will usually out-perform a leaky building with fancy equipment.  Make field-verified infiltration tests the default for your projects.  Make your designers trace the location and detailing of the air-weather barrier all the way around the building drawings in both plan and section. There's often one of those "And then a miracle occurs" moments where walls meet roofs, etc.
• Materials: Develop a reduced palette of the 'typical' finish materials your firm likes to use, screen them for the health-impacting nasties or high global warming potential. 'Declare' labels make this pretty easy these days. Then tell your designers that they can pick things outside that palette only if they (not you) prove that their proposed choice isn't nasty.  If your contractor wants to substitute a material that's got endocrine disruptors, tell them (if they are male) that it will make their penis smaller.  That usually gets their attention.
• Heating & cooling equipment.  I'm not sure what the question is, but the answer is usually a heat pump. For hot water too. Learn to say 'Heat Recovery Chiller' three times very fast.  Challenge your engineer to get over combustion, already.  
• Lighting.  Tell your lighting engineer at project kickoff that they have to submit the Lighting Power Density budget for the project at SD and each phase thereafter, and the starting point should be 1/2 of what ASHRAE 90.1-2016 says is the maximum allowed.  They will squeal about IESNA goals, which are actually huge ranges, and horizontal illuminance levels  that tend to be a poor measure of lighting quality. Give people a $100 light meter and have them walk around to the spaces they want to work in and record the actual light levels. Dare to be dim. The goal is to provide all the light people need, but only where and when they need it.
• Appliances. If you do residential projects and do everything else right, appliances will be more than half your energy consumpion, so it's unglamorous work but you need to obsess about appliances. There's probably somebody in your firm who actually gets excited looking at appliances; release the Kraken. 'EnergyStar' is a starting point, not the finish line.

I love a good simulation with appealing color graphics.  I think they can sometimes be a trap.

Oh my gosh, Z...this should be an epilogue to the AIA's "An Architects Guide to Integrating Energy Modeling in the Design Process"! This is good stuff.

Thanks for this Z. Could I share your best practices with our internal Sustainability Working Group if I give you credit?

Related to Kristian's comments on Cove Tool and its inaccuracy on end use splits, I believe I've mentioned to this group that in an early test of the tool compared to a verified project energy model we found the relative heating and cooling annual end uses to be very different. In discussion with the CT folks, we found that the model assumes just one thermal zone per floor, which is particularly problematic for larger buildings in heating climates, where a typical building would have perimeter heating, and interior zone cooling. There may be ways now for correcting for this - specifying / creating multiple thermal zones - I'm not sure.

What we have found most useful about Sefaira and cove.tool is energy literacy among staff. If they click the buttons a bit and take some training they get the terminology of what is important and some basics on how design has an effect on energy use. (ie peak loads, zones, effects of orientation, seasonality, energy end use pie charts for different climates, window to wall ratio). This allows them to coordinate third party energy modeling more successfully and conversationally. -Kjell

FWIW, what Cove Tool and Sefaira offer is called 'reduced order modeling.' From a recent DOE article: "Understanding that early-stage design requires rapid, iterative feedback and has few HVAC system details, ... an approach based on reduced order modeling (ROM)." It now seems like we are beginning to make a smoother transition from ROM to more detailed energy models, which can be important for continuity in an integrative process. Here is a link to the full article by DOE which offers further insight: Cove.Tool Officiates Perfect Marriage Between Reduced Order Modeling and OpenStudio | Department of Energy (if the link is disabled, google the terms in the line above) Ramana Koti Senior Associate, BEMP, LEED Fellow D 404 253 1478 Responsive Design LordAeckSargent.com

To Jason and Doug's kind reaction to my screed... Thanks!  and Doug, please feel free to re-use the material (and please correct the typos I see crept in during my late-night post-Ida stress relief typing). To follow up on Patrick's hat tip to Paths to Carbon Zero... We were able to produce that work through a pretty unusual alignment of the stars, where we had one staff member (Waleed AlGhamdi) who was trained as an architect but had worked for years doing nothing but energy modeling for a consulting firm before joining EDR, and two other research fellows (Kevin He [3 months] and Kelsey Wotila [12 months]) working full-time on carbon accounting.  Not every firm is in a position to support that level of investment.  And holding on to people with that level of skill can be a challenge; Waleed was called back home to be with his family in Saudi Arabia (where he's now heading sustainability efforts for a company charged with building 400,000 housing units), while we were fortunate enough to convince Kelsey to stay on as general staff, to help her gain experience across all the aspects of architectural practice and earn her license.  Before Waleed, we were lucky enough to have Jake Dunn on staff (who brought years of modeling experience from the Integrated Design Lab in Boise and is now back in the Pacific Northwest at ZGF where he's an Associate Principal, working with Chris Chatto), and before that Corey Squire (who went on to Lake | Flato and will soon start at Bora). Corey and I share the passion for the 'best-practice practice' approach but I've learned a ton from the rigor of the analyses done by Jake, Waleed, Kevin, and Kelsey.    We're just crazy lucky here in New Orleans, no matter what the National Weather Service says.  To Ramana's observation about reduced order modeling (ROM) vs detailed modeling:  I think the place of deep simulation is to put gut feelings to the test. That old saying applies: “It ain’t so much men’s ignorance that does the harm as their knowing so many things that ain’t so.”  Example: We're in the middle of a lab project where it took some pretty sophisticated energy modeling to come to the conclusion that the HVAC system that "everybody knows" will be the most efficient (DOAS+chilled beams) in fact was not for this project, given this project's program mix and and climate.  I doubt rules of thumb could have got us there.   But the differences in energy use between the HVAC options we modeled was tiny compared to the savings from working with the client to change number and type of fume hoods--and sophisticated modeling only served to confirm the simple online calculators for the savings.    So, I think we need both "best-practice practice" to avoid the stupid stuff, and sophisticated modeling to settle the subtle questions.  Maybe 'reduced order modeling' has a role to play in between these two.

Patrick, I initially typed up a response a few days ago and it magically disappeared before I could post it. Since then conversation is more than I have to offer, but here ya go.... I was new to this 4 years or so ago and I have learned a lot from this forum but we have decided to keep it really simple. So what we ended up with are a couple different approaches.  1. First step/attempt was to build understanding and capacity. After testing out Sefaira and then Cove, we settled on Insight. We created a simple workflow, and train designers as needed as soon as they start massing a project. For the most part this has caught on and while we know this is low tech modeling, like one responder said, we have at minimum gotten people fluent with the concept of a pEUI and how to do quick models, pull the levers and understand what is going to influence your project- it is maybe 1/2 day committment. By the time the energy modeler is involved our team is actually engaged in the conversation and not just looking at the work being done in the silo. Because Insight is simple enough, the only time we do the work for the team is when the designer doesn't know Revit, there are still a few here. 2. For those projects who need a little more but don't have MEP or an energy modeler on the team, we have an on-call energy modeler. We have 2-3 who have signed contracts with us and any design team can tap into them for quick modeling or even advice and then the office pays them for the time as needed. Ideally their fees are "funded" by the utility incentive money our teams get from their projects which is usually $5000-$15000 per project. If the client is willing to contract them directly we do this too. But the idea was to eliminate the barrier we often heard "the client doesn't want to pay for this, or we don't have a modeler on board yet". This has worked well for a couple of projects, but we are just getting this off the ground and not all design teams know about it yet. 3. We have a High Performance Analyst that supports all of our design teams with Rhino/Grasshopper. We brought someone in as test pre-covid and the principals saw the benefit so we just hired a permanent person this summer. I manage his time, and we help teams perform early studies to test out designs anywhere from daylighting, to solar radiation, sun shading, massing and shadow studies. It took us about 16 months or so to build up a menu of studies we could do and then test them out and refine them. He is also busy doing daylighting studies for our interiors group and doing the LEED and WELL docs. Part of his responsibility is to share on bi-monthly basis with our internal Grasshopper user group, basically anyone who wants to learn, and coach them on how to do the studies on their own projects. Anytime we engage with a project team our first preference is that we coach someone on the team to do the study, but this has proven to be challenging as the teams are understaffed and most people don't know the software. But also the advantage is the lessons learned and ability for us to get better at it and be consistent. It also allows me to be in touch with more projects so I have a better sense what is happening across the firm. But the goal is definitely to build capacity and redundancy so the knowledge is shared. Maybe in a year I will have a better balance of this work load.  Hope that helps. Kristen

Hi Patrick, Most of our teams work in sketch up for early work, then there are a few that use Revit massing, and then even fewer who use ACAD. So we tried Sefaira out for almost two years, we were paying a lot for not a lot of use w/ inflexible terms. I was not part of the group that tested out Insight and decided it was preferable for our office, but ultimately I think w/ all of the software, we found people most likely had to spend a little time cleaning up a model or rebuilding a massing model, and in Revit that was quick and easy to do. So no matter the state of the design or level of detail, we build a Revit massing model (or try to get the team to do it) and then depending on the intensity of the study that is needed we either keep it simple, or develop more detail on the model per windows etc. – the latter happens less often. Most often the projects that engage us are the lab or office core/shell buildings and multifamily or student housing projects and then a handful of other building types. I would think all of these would be applicable to k-12 programming. We most often use Grasshopper for doing: 1. radiation analysis – looking at building orientation, solar pv design, shading devices, building form –informative to the design team, but also helpful in having conversations w/ client/owner about location of roof terraces, outdoor space, balconies, shading devices, wwr density per façade, and pv locations, also used a lot in public meetings or approval process documentation to share w/ city jurisdictions who are pushing us to make sure we are looking at all of these considerations. 2. sunlight hours or shadow study – mostly the shadow studies are required by local jurisdictions, but we also use them to show sunlight hours on the surrounding site to help the design/team owner locate open space on the ground and on roof terraces, or for studying the impact on the surrounding area. We do this on a lot of our master plan work as well. 3. Thermal comfort – we use this less often, but look at sunlight hours and then comfort to determine when temps might make inhabitants feel too hot or cold – we have used this to study amenity and gathering spaces, atrium/lobby spaces etc. 4. Daylight autonomy and daylight factor to study daylighting and glare. We have used it mostly on office and multifamily spaces to comply w/ LEED or WELL credits, and then also to study placement of skylights to balance the daylighting coming from windows on the opposite side of a room. We have also used this to help office planning teams realize where hot spots may be on their layouts so they can plan amenity vs. desk spaces. And then we have used it to study the size of windows in a multi-family project to convince developers they will get adequate daylighting (and views) from smaller windows. 5. We do often use the weather data and show a locations wind rose, dry bulb temp, sun path diagram/temp, and universal thermal index, etc… Kristen Fritsch AIA LEED AP BD+C WELL AP Sustainability Coordinator ELKUS MANFREDI ARCHITECTS [tel] 617.695.7954 [email] kfritsch@elkus-manfredi.com WFH# 206-356-5296 From: Patrick

Hi All, Thanks for such great responses. A lot of useful discussion here. Thanks Z for the great list.     Kristian, I share a lot of the same frustrations with the low-resolution energy modeling software options. We tried Sefaira and actually used it a lot in the past, but since they were bought up by Trimble their customer support has gone downhill and it is VERY hard to get Revit models to upload and work smoothly without a lot of painful back and forth with their elusive support staff. That is a primary reason we felt we had to move away from them. Cove.tool is nice in that you don't need closed rooms and spaces for a basic energy model--certainly doesn't make it super accurate, but at least it's fast for a quick check-in at each design phase. And they are very responsive to help requests. We tried out Insight too, but the lack of ability to specify basic properties of HVAC and building envelope seemed too limiting.   I agree with Kjell though, I think the number one reason we like cove.tool is really that it increases staff literacy about energy performance. Because we make each project team at least try to start the cove.tool model on their own, they have to wrap their head around pEUI and how it's affected by different parameters. They play with the insulation, HVAC, and glazing properties to get a sense of the sensitivities. And, more importantly, it's the only way I can think of to effectively model all our projects for 2030, since we aim to have 100% of projects with a pEUI. It's easy to integrate with our Revit workflow. I agree that it is hard to use cove.tool to actually make effective design massing decisions because the differences between options are indeed often not as big as you would hope. However, we have used cove.tool to compare different glass and insulation options, as well as wall to window ratios, and it has been useful as justification for going for a more expensive glass option, for instance.  The optimization tool has never really been useful to us either. The cost info just isn't accurate. I end up just fiddling with individual inputs to compare glazing and insulation options. For instance,attached is an energy measures vs EUI analysis we made for a recent project using cove tool toggles. Could we have gotten to the same place by just following best practices? Yes, likely. But this helps clarify priorities a little better. How accurate it is though is anyone's guess...     RIght now I am the only one using Rhino/Grasshopper/Ladybug. We've used it mostly for daylighting analysis. Our work is all in Revit and conversion can be a bit of a pain. I would also love to hear more about what others have found/achieved through the more sophisticated Ladybug tools. Excited to keep the conversation going, Misha  

Hi Patrick - Seems like most of the responses here are focused on the energy modeling and daylighting, so I thought I'd jump in on the LCA side of things.  As Kjell mentioned, I help run LMN's WBLCA process.  We primarily use Tally, although we do some material selection studies through EC3.  With Tally, our original thought was to train up one or two experts and drop them into projects when the studies needed to be run.  This proved to be inefficient.  Tally is a relatively simple program to pick up, and the difficulty of the study is often related to the specifics of the project - how it's modeled, what the material assumptions are, etc.  We switched gears and started a process where each project team has a member that gets trained up on Tally and they maintain the model definitions over the course of the project.  The initial training takes about 30 minutes, and then I do check-ins over the course of the modeling process to discuss any issues they run into and any Tally quirks that they need to be aware of.  So the set up is similar to Zs - We have one person who's "the expert" - someone who's used to looking at models, knows the common pitfalls, and is there to provide guidance on the modeling and the results - and a lot of individuals who may not have the time to be LCA experts, but understand Tally and are able to ensure that study is accurate to the design.    In terms of time for all of this, training project members is relatively quick and easy.  The best performing project teams have set aside 15 minutes at the end of the work week to review the material definitions and to add anything new.  For the office expert, the time commitment is larger, and it is helpful to have time built into the workflow for training, troubleshooting, developing guides and QA/QC tools, and ultimately distilling the LCA results.  However, even then, I find that I'm working on it only a few hours a month.   Hope this helps, and I'm happy to discuss the process further.     -Justin

Patrick, I should clarify that with Tally, while I don't believe you need to have an expert, the program still requires some understanding that comes with practice.  It's a deceptively simple program to learn - It's loaded within your model and then you apply materials information based off of what your assemblies.  Once you've labeled all your materials, you run the study, and you get your results.     What is difficult is learning how to work the way that Tally wants you to work.  Certain information doesn't want to report correctly -- most families and occasionally curtain wall panels and mullions. Tally also has a round about way of entering information for certain materials -- doors for instance are included as single "door" materials instead of glass and steel or any other material that they are composed of.  Understanding Tally's way of reading information from Revit is another issue that catches people off guard at first -- is it reading the length of a wall or the area of a wall, and how this applies to the data coming out.  None of these are difficult to deal with, it's just that the Tally system can be somewhat opaque for the first few modeling attempts.     

I saw Z's list when it was posted, read some of it, knew it would be good and sent on to my office. Then I went back and read it carefully and was blown away, but not surprised at how brillant it was. Right up there with the 12 Product Rules that Building Green posted a few years back. Nobody says it like Z.