Sustainable Data Centers and Embodied Carbon

Increasingly, data center owners are looking beyond the environmental footprint of their operations and looking at the embodied carbon—the carbon footprint from manufacturing and transporting construction materials—of their facilities. With green commitments taking hold for them and their own customers, there is a need for collaboration among owners, designers, consultants, general contractors and manufacturers to deliver new, lower carbon approaches.

In this installment of the roundtable discussion, panelists touch upon embodied carbon, the roadblocks around alignment in this space and where new methods can be used to push forward.


In Part 3 of the Strategies for Sustainable Data Centers Roundtable series, DPR Construction, Equinix, Brightworks Sustainability and Sheehan Nagle Hartray Architects address strategies for embodied carbon. Watch the video.

The roundtable discussion panelists included (from left to right in the video):

Ryan Poole

Ryan Poole

Global Sustainability Leader at DPR Construction

Greg Metcalf

Greg Metcalf

Senior Director, Global Design at Equinix

Josh Hatch

Joshua Hatch

Principal at Brightworks Sustainability

Denis Blanc

Denis Blanc

Director of Sustainability at Sheehan Nagle Hartray Architects

The roundtable series will be shared in four parts and will include segments covering energy, water, embodied carbon and what's next.



This discussion has been edited for clarity.

What's top of mind in the data center space for energy, water, embodied carbon—for transitioning the built environment into being a more sustainable and resilient resource?

Ryan: One of the last topics we're going to talk about and dive deep into is embodied carbon and how embodied carbon is tied not only to owner's commitments around the greenhouse gases and how that divulges into scope three, how it's also tied to the responsibility of the design community, consultancy community, the contractor community, and the manufacturing community. There's this collaboration approach that has to happen for us to drive innovation forward. We're going to dive deeper into some of the roadblocks that we currently have around alignment in this space, and where we're using new means and methods to drive forward.

[0:50 Watch from here] Denis, I'll start with you. Thinking about when we're doing this upfront planning, all these things that are driving—particularly from a structural standpoint—lowering embodied carbon within data center halls and the admin spaces.

Denis: Indeed, Ryan. That's really on our list of things to investigate, all that you mentioned. We're trying to look at this in terms of prioritizing what's really important. So we can really use our effort on what matters the most in the overall profile, for a data center carbon profile.

We know a few things. We talked about how the systems, building systems, MEP, electrical distributions are really heavy due to all the metals that they use. We know that structure envelope are big things to look at, at the very early stage. Indeed planning has an impact. How do we plan our buildings on the campus? How do we minimize the footprint? We are looking at all of that. What we're trying to do with CLT is to really understand what it means in terms of implications for fire requirements. There may be obstacles but it's something that we definitely want to look at.

From what I've seen so far with embodied carbon, the accounting and the equation is simple but there is a lot in the carbon intensity of a specific product of our system that comes into play to determine what the impact is for a specific system. I've seen studies that would show that a steel structure may be actually less carbon intensive than a CLT, so it depends on the location. There's not one simple answer but we know that obviously, concrete is a big thing. There is a lot of research that is being done, not necessarily by us, but by people before us on increasing data strength criteria for concrete, which makes it less carbon incentive, in terms of the design. Foundations obviously have a big impact. So, does that imply that we could choose between two sides? One that has a good, sword bearing capacity to avoid having to spend a lot and build huge foundations, envelope CLT. All those things we look at. We talked about the tools that we have at our disposal to evaluate all of that. This is in progress. This is not an easy, immediate output. Right now there's a lot of post processing from extracting quantities from Revit models and bringing them into a lifecycle assessment software. There's a lot to understand and to make sure we are actually quantifying things correctly. With carbon, a lot comes into play but that's something that everybody wants to look at.

Ryan: It's more recently that we've continued to discover the larger impact that embodied carbon has. Looking at the raw extraction of those materials, the processing of them, how they get to the plant, and how they're manufactured when they're there. The reality is, we still have a long way to go in driving that transparency and clarity around that. We have EPDs that have given us some standardization but there's still a long way to go in making sure that we're actually consistent in the way that we're running those calculations through those environmental product declarations. DPR's supporting research with FSC right now to to try to make sure that we're actually accounting for the right type of carbon sequestration in properly managed forests. That can be different depending on how the forest is managed.

I think you made another great point, Denis, is there's not one solution for every application. Where the exploration of recycled glass poslin as a replacement for cement. There's only a few of those plants throughout the world, so depending on where you are located to them, your embodied carbon for utilization for that could be significantly different than if you try to ship it. I think it's the same way with wood.

[5:50 Watch from here] We have to balance out that transportation aspect of the resource that we're trying to use versus the sequestration aspects that's built into it from a lifecycle perspective. There's always this fine balance and it's going to obviously take a lot more collaboration from all of us to continue to create that synergy and the demand and push that's got to come from the manufacturing side.

Josh: In the material space, the green building community started out with some kind of low hanging fruit. They used materials that have more recycled content that are locally provided. Choose FSC wood over non-FSC wood. I think a lot of those choices were like-for-like substitutions. It's like you have an FSC 2"x4" and you have a non-FSC 2"x4" and only real question is, is it available and how much more does it cost?

I think when we're getting into these more complex topics like embodied carbon while there are similar choices. Drywall that has a lower carbon content than other drywall or carpet that's better than that other carpet. The bigger opportunities are in structure and design. You can't just swap out one product for another, you have to rethink how you're designing the whole project. In that big process of rethinking, such as "is a structural system heavy timber, steel, or concrete panels?" It's a big decision from a variety of perspectives and involves much more teamwork to make the right choice. You have to have your structural engineer being thoughtful about how you build the project up in different ways: from a structural standpoint, from an architecture standpoint, what are the implications, and so on. There are still a lot of things that happen in data centers besides just holding the building up. Cables have to go places. Servers have to go places. Then, from a construction and price standpoint. Things have different costs depending on market and supply chains that are available there.

We're really rethinking about how we design and build buildings from the body carbon standpoint. We can make the easy choices. I think a lot of the foundation we're trying to build on right now is with EPDs, with getting transparent data out there so people can make informed choices. I think that's going to drive a lot of the changes that we can see on like-for-like substitutions. I think we're trying to build the infrastructure to make the bigger impact choices. How do we rethink how we design buildings now that they're becoming more efficient and more supplied by renewable energy? Embodied carbon's becoming the remaining big thing we need to address. Unlike operational carbon, which is going to look better over time as a grid gets green, for embodied carbon when you build a building most of it is spent the year you build a building. The products that impact is all spent the first year. Whereas operational, we have more time over the life of a building to let the grid green.

Ryan: You make a great point that you can't just look at the GWP—the embodied carbon that's associated with it—and say that's just the better choice because of that aspect of it. It's a lot more holistic in that and looking at the design implications. They have different properties depending on what type of material you use and then that can affect the design efficiency.

[9:00 Watch from here] You have to go back to your energy model and looking at that the r value that you get from insulation the “do you have to add more material to it to make it more fire resistant?” those types of things. That can play a heavy weight into why it's so important for us to collaborate early around supply chain impacts, where the materials and resources are located, and what's constructible. What's achievable and what's not?

Denis: I'd like to add things that we do at SNHA, which is at the scale of the product development. Sometimes we do research for new design on certain components such as the hot aisle containment system. It's not necessarily initially demanded for it's sustainability attributes by a customer, but it can be for cost and time saving. However, we can also add the layer of sustainability and everybody will be receptive to that. It goes with trying to design for easy disassembly. For example, reducing the number of components. One example would be a hot high contentment system which in one case uses aluminum framing, in other cases does not. You have a direct reduction of material use which has good impact on lowering embodied carbon. We think about this level of planning, building footprint, major systems, but also it can go to smaller systems that are very present in the data center. Which goes also with eventually modularity, ability to expand, easier disassembly and offset upside options.

Ryan: Two aspects there Denis that are great are: the speed to market (actually getting the building up) and how quickly you can do that. Some of the various solutions that we're talking about, particularly around pre-fabrication, make that a more rapid process. People don't think about it but there's financial gain in that and also a lot of embodied carbon savings from a reduction in construction activities on the job site. That's one of those return on investment aspects that we have to look at. Are you able to get the data center space open faster from utilizing pre-fabricated processes?

The other things that we're looking at is how we actually phase some of the development of that. Longer cure times allow for less usage of the amount of limestone, you can have higher fly ash contents, and things like that which can also drive down the embodied carbon within concrete. It's looking at the actual phasing and saying "can we allow more time for curing of slabs and such?" and phasing it out. That has, just from a logistics standpoint, reduction in embodied carbon. That's where there's super important collaboration to think about, that phasing from a construction design and owner's perspective of when things need to be turned over.

[12:25 Watch from here] Greg the thing that's coming to mind for me is where are owners looking at this? How is it important to you the connection for embodied carbon that feeds into how your commitments around your greenhouse gas protocol, getting into scope three and how you're thinking about analyzing these solutions with the team and making those decisions as you get into and past the design phase?

Greg: It starts with benchmarking and conducting whole building lifecycle analysis on what we are doing today to give us better indication of what we should do in the future. Considering components of a building and in a choice of A versus B, steel versus concrete, they're the kinds of decisions that we're actively looking at in new construction with regard embodied carbon. With good data and assistance of partners doing those studies we can make better choices, and overall coming towards a whole-building LCA number that we can believe in.

Ryan: There's certainly lots of mechanisms that are helping us push towards that. Paris Climate Accord and having companies sign onto that, science-based target initiatives and some other frameworks that we have. There's also a lot of us that are taking more aggressive steps on our own and going a little bit above and beyond, which I think most of the folks here have.

Josh: There's been a lot of work for many years in looking into embodied carbon, but really it happened in the last five years that it's gained popular use. I think one of the big reasons of that was that USGBC LEED, when it came out with version four, the shift of materials credits to just being focused on recycled content, regional materials FSC and low VOC products from a health standpoint. They started saying we also need a more full assessment of the impact of materials, the EPD and the HPD disclosures of environmental and health impacts of the materials that we're using all the way through their lifecycle. At the time there wasn't that many of them out there. There were some progressive manufacturers that had done that work but many had not yet because there wasn't as much of a demand for it. By putting that stake in the ground, for the first year there wasn't much either. Two years out, three years out, all of a sudden there's been a rush to the market of people disclosing what's in their products. That's getting to be better and more comparable data. On top of those disclosures, we've been able to build tools to help solve climate change.

There's still a lot of gaps in the data sets. For the data center industry in particular, we're working on the MEP systems and the IT systems, some of the more focused systems, to get that data on. We're catching up quickly there. It wasn't a credit that companies went after until the last three or four years for the most part. Now it's been a rush with just the increasing awareness and pressure to make our planet a more livable planet for a longer period of time. I think it's shifting from something that you do voluntarily, because you're progressive, or because you feel like it's the right thing to do to something that you have to disclose from a business as usual perspective. I think that shift not only is helping all companies understand it and consider it but it's also raising the stakes of that disclosure.

When you're voluntarily reporting something, you might be willing to do it at a certain level of precision when it's being provided to your shareholders in a reporting that goes through a regulatory body. There's a much higher emphasis on making sure it's to industry standards. There's a lot of work that needs to be done in those industry standards, just getting comparable data between products, making sure products all have data and the methodologies that are being used. There's methodologies we use in design to make better choices. There's a separate set of methodologies that are being used at the ESG reporting level. We're trying to connect those two so that we're not doing redundant work, so that the work we're doing on designing buildings that are higher performance with lower embodied carbon is work we can hand over to the ESG department. Similarly, it's work that Equinix can then hand to their customers when their customers are asking for their reporting because one company's scope three is another company's scope two and one. It's all related. It helps to have companies take leadership on science-based targets and ask their vendors and their supply chain to do it because it then gives the business reason for those companies to do it.

Ryan: There are actually governmental requirements now that are tying us to some of this disclosure. The really awesome thing I think for all of us to sit back and reflect on a little bit is it's doing its job. It's doing exactly what it did when we started to have more transparency behind operational carbon, which was to get people to focus on reduction and to get people to create something better because we now have visibility into what the impact is. It's really awesome to see that transparency that we're starting to drive in the manufacturing space actually create better products.

Greg: Josh mentioned earlier, but it bears repeating that VPPAs have been the industry's method of procuring green power. The number you quoted earlier of how much green power has come on the grid and what percentage of it is as a result of data centers buying it is an example of what a collective buying power has done to change the electricity system and stimulating it in the right direction. We can, through mechanisms like LEED and other mechanisms, drive better buying practices that are going to change the industry's direction over time.

[18:05 Watch from here] It's good to bring that back into play of how that fits into the embodied carbon space. There are these numerous initiatives and things, but we've got to get more collective and synergized around how we're actually approaching that.

Josh: I think PUE a great example of everyone was starting to calculate it, but it took a white paper to say here's the approach we should all be using. That got a lot more companies doing it. They got some consistency in approach, not a hundred percent consistency, but it provided a lot of transparency to the process. Then companies were able to compare. The same thing is starting to happen, and something we're working on right now, is to get a bunch of companies to disclose the embodied carbon work that they've done on their projects. A lot of the work's being done internally to influence decisions and that's great, but the more that we can put it out there so that we can compare it not from a which is better or who did it better but from how aligned are we? What are the different methods? How do we better align those methods so they're more comparable? How do we start to understand what is the the minimum standard?

These standards are being created by people testing things out. These are new methods. The more that we can get out there and see what other people are doing, it helps us collectively understand what's good, what's bad, and what we need to get better at so that we can move forward and have clarity. For regulatory, green bonds, all these entities that are trying to support the same stuff we're doing, they don't know what they're asking for yet. They're asking for disclosure for things that don't already have methods necessarily or they're asking for the value of having consistency and transparency in these new emerging analyses so that we can get better at it.

It's also market demand. Data centers have been huge in bringing new renewable energy onto the grid. They can be huge in helping the embodied carbon movement take off. The more demand we get for lower carbon products, manufacturers are then going to have to study what they're doing. Some of that is their energy purchases, so that creates demand for more renewable energy buying. It creates demand for better manufacturing processes. Those might not be industries that have as strong as a business case for sustainability relative to more consumer facing brands. I think every sector has a different business case for how fast and why they're going towards sustainability. The more that we can make goals that trickle down to the companies we work with and partner with them to help make it happen. I'm going to ask you to set out and put a science-based target out, but I'm going to ask you that because I want to make progress on my own. I'm going to help you there. If you commit to making reductions, I will commit to buying your product. We need those types of partnerships so that manufacturers can make the big steps and be rewarded for it.

Ryan: Yes, and it's where we all need good alignment in a framework which is where rating systems come into play. We absolutely need those to continue to evolve, to keep up with current times and what's most important for a particular market space. I think it's important to talk about the consistency in the framework, the validation that you get on those.


[21:16 Watch from here] Josh, maybe from your perspective because in the consultancy world that's a big piece of what you do. Let's talk a little bit about the importance behind those rating systems and providing that data collection and analyzation and framework that supports our client's goals.

Josh: I think rating systems have played a huge benefit in the market in providing clarity and consistency, rigor and credibility to what we're all trying to do. Before USGBC published LEED, anyone could claim they had a "green building" because they did X, or they did Y, or they did Z. It was hard to compare. Is this a green building because they used FSC wood, is this a green building because they're more efficient than code? How much do you need to do to get to a point where you can call yourself a green building? It's really nice to have a third party that that says here's a checklist.

Getting consistency on what you need to do, getting it validated by an external party so it's not subject to claims of greenwashing or stating what's good and ignoring what's bad helps bring everyone to the same level. Those systems have to be designed really thoughtfully and I think they apply to many different building types. They're trying to create a “one size fits all” and I think data centers in particular haven’t fit as well into some of these rating systems because they’re such unique projects. The data center version of the LEED rating system came out. That’s been really helpful, but there’s still more work that needs to be done in terms of appropriately targeting impacts with the effort.

Greg: I think it's worth reminding everybody that data centers are somewhat unique maybe similar to some industrial processes, that the operational carbon vastly outweighs the embodied carbon through the life cycle of the building compared to an office-based project, those two the embodied carbon significantly outweighs the operational carbon in most cases.

[23:09 Watch from here] We also have to think about from the owner's perspective of transitioning of that space. New equipment being brought in, reinvigorating the space, making it more efficient as time goes on. That’s also embodied carbon that’s packed in throughout its lifecycle as well. Thinking about the longevity of it and when certain components have to be replaced and things like that are upgraded.

Greg: Yes, we definitely think about the longevity of MEP systems and are selecting MEP systems for longevity to avoid replacing them. An interesting debate point is around battery systems. Battery systems, lead acid systems, might need replacing on a five- or six-year interval. They're highly recyclable. The plastics are highly recyclable, the lead is highly recyclable. They're perhaps the most recycled item in the world. Lithium-ion systems on the other hand, can last 25 years in some cases. You'll get a degradation of capacity over time, but they can last much much longer. Studies on those types of systems are things that we're actively looking at.

Ryan: And increasing the recyclability of those like you mentioned. Lithium has its own challenges. Continuing to evolve the recycling processes that we have so that you can have more and more access to that creates really a circular economy in the way that we're thinking about how we transition this old equipment back into new equipment through the extraction of the material.

I think an interesting aspect that's sometimes not thought about with embodied carbon are the actual construction practices that happen on the job site. I mean, that is part of the embodied carbon that it took to create that building. We've done a lot with our progressive data center clients around tracking those metrics. What are our greenhouse gas usage from fuel expenditure when we're on a job site creating it? What types of tools and equipment are we using? How much water did we use? Getting metrics around that, and we're starting to get much more sophisticated around the automation of that data collection. Which is working closely with our data center clients to say we need smart meters that are out here so we don't have the inefficiency of people going and reading that information. We need that to feed into a platform so that we can all sit down and have access to that information. It helps drive those better design choices, or maybe even different construction means and methods along the way.

Greg: That's an analogous problem to what we have in our existing data center fleet of going out and enabling it for the metering that we can report on. It's not easy in existing buildings and it requires investment, but we are going and doing that so that we can improve PUE across the board.

Josh: You spoke earlier about modular and construction pre-manufacturing. I think that's one of those triple wins. It speeds up delivery. When you're making something in a factory, there's higher quality control and less waste. That means less re-do, less overages of materials. All those things also contribute to reducing embodied carbon in ways that aren't immediately apparent. I think we always think of buy product B instead of product A, without thinking about if we make product A perfect every time and don't have to cut off the end of a piece of wood to make it, those are all savings too.

[26:23 Watch from here] You're hitting in to in construction practices, Josh, one of the things we focus on most are lean practice principles. How do we actually plan waste out of our process, making sure we have proper planning techniques with using virtual design and construction for clash detection, avoiding rework through those practices.

Ryan: Making sure, just like you said, if we're cutting a piece of drywall and instead of cutting a small piece here, have we actually planned it out so that we can use the most of that piece of material before putting it into a waste stream? Then, continuously thinking more and more about the next phase of that waste. If we actually have construction waste, how did we select the right materials upfront to make sure that they're recyclable and that they can go back into a circular economy? Then how do we partner with the right vendors to have extremely high diversion rates throughout that process? That's another really big part of the embodied carbon component on data center spaces that we've got to continuously improve.

Josh: That's a good point on how waste, embodied carbon, and construction processes like these are all interrelated issues. We have clients that have zero waste initiatives and zero carbon initiatives. They're not independent initiatives. They support each other. It takes an architect, a contractor, an owner, and certain consultants to help them through the process.

Back to the certifications landscape, there are disclosures for products, there are building certifications, and there are corporate disclosures. Helping those work together is a process we're knitting together real-time because these weren't present in the market 10 years ago. The more that information can be passed along and initiatives can support each other, the better. Also, helping align what requests your customers are asking of you. Often, we talk to sustainability people that are within companies helping respond to customer requests or regulatory requests. They spend much of their job repurposing the same data to answer customers that have asked the questions a different way instead of focusing on innovation to actually make the performance better. The more we can align and share, the better we can make progress together and make it easier on ourselves.

Greg: I think you've struck on a key point there: alignment amongst people requesting information about the format that they requested in. The unique configurations that people ask for the data in when they ask for green sustainable reporting creates a huge amount of work to answer those. We'd be much better focusing our effort on having a standardized reporting methodology that would allow us to put our effort into other areas and concentrate on that. The reporting aspects of sustainability have gotten pretty painful without a standardized methodology.

Josh: There's a gap and that took a lot of years to create, but it's been really effective in the financial world. We need a carbon version of gap. We need a really consistent, clear way that we're assessed so that everyone knows the level of precision and the method that needs to be done. Then we can spend the time not doing the accounting but doing the work.

Ryan: There's a great correlation here. We're talking about driving demand for high recyclable, low embodied carbon content. The material creation is a similar comparative, it makes it more economical at scale. I think that's similar to the process of us actually doing the analytics. The more we do it, the more economical it becomes for clients to be able to access that, the more and more people that are buying into it because we have more efficient platforms. We've gotten more efficient in our collaboration methodology. We now have access to more data of the supply chain, so the process continuously become more and more efficient.

Sustainable Data Centers

Data centers have become a cornerstone of modern life. At the same time, these facilities have traditionally had significant environmental footprints. The question is: how do we support the growth of digital infrastructure while also better managing its energy, water and carbon footprint?