Decarbonizing the MIT Campus Final report for 4.s42 – Spring 2024 Christoph Reinhart (Instructor) Akshata Atre Elizabeth Bernhardt Meghan Blumstein Ippolyti Dellatolas Kiley Feickert (Teaching Assistant) Emile Germonpre Zoe Le Hong Ali Irani Yewon Ji Nikita Klimenko Amanda Kirkeby Jazmin Mucino Sanjana Paul Alison Maguire Morgan Johnson Quamina Nada Tarkhan Ben Taube Graham Turk Ray Wang Sam Wolk
Executive Summary Scope 1 and 2 greenhouse gas (GHG) emission from MIT owned buildings in Cambridge have largely remained constant over the past decade. The climate crisis demands that MIT takes a comprehensive set of measures to reduce its on-site carbon emissions to zero as soon as possible. To realize this goal, a series of integrated technology updates must be initiated: Figure 1: MIT campus section with key upgrade strategies - Comprehensive building retrofits across campus would reduce carbon emission by up to 40% and add to occupant comfort and well-being (1). - Adding PV on all suitable rooftops would decrease our emissions by an additional 5% depending on the carbon content of the grid (2). - Electrify our district heating and cooling system by continuing to change from steam and chilled water distribution loops to either hot and chilled water loops or a single ambient temperature water loop (3). The system choice has wide ranging repercussions and needs to be carefully studied. Both system types would be connected to a 300m deep borehole field to act as a seasonal heat source or sink (4). The hot/chilled water solution would be centrally conditioned via industrial-sized water-based heat pumps. An ambient loop would reduce energy use when coincident heating and cooling occurs but requires individual heat pumps in all campus buildings. An example for this approach will be the new Metropolitan Warehouse. A distributed system would require more space in connected buildings and might be more disruptive to install and maintain but is better suited for a phased approach. - In any case, the district energy system needs a clean source of electricity, for which we recommend pursuing two emerging emissions-free energy sources with strong ties to ongoing MIT research: deep geothermal boreholes (5) nuclear batteries (6). Alternatively, grid electricity can be used, assuming the grid itself is decarbonized and has the spare capacity to take on the campus loads. - While carbon capture technologies are evolving quickly and could largely remediate remaining emissions from on-site combustion, it remains unclear how and where MIT could sequester the resulting CO2 (7). - Given current electric tariffs in New England, battery energy storage is currently neither economically viable nor technically required since our diesel generator suite can provide emergency backup for resiliency. Economic boundary conditions and grid stability may change as the grid decarbonizes, in which case energy storage technology will become a key
technology for us to consider for any scenario that heavily relies on the grid. While future electricity loads from MIT employees’ charging their electric vehicles during the day will require extensive additional capacity, this component is ignored in this report since it focusses on scope 1 and 2 emissions. It should be noted, however, that the batteries of these EVs offer significant local storage potential for demand-side management if MIT were allowed to temporarily discharge them if and when needed. Lastly, thermal batteries are an emerging technology that may hold promise for campus decarbonization goals but need further evaluation. To implement a net zero reduction pathway for the MIT campus in a timely manner, we recommend that MIT takes the following measures immediately: 1. Form a Building Energy team that maintains standardized retrofit packages for different building types, manages a campus-wide retrofit plan until 2050, coordinates with individual project teams and verifies post-retrofit savings. To elevate the status of the group, it could be housed within Facilities, work with MITOS and a standing committee of faculty experts and report annual GHG emissions to senior leadership. 2. Understanding the ground below MIT is crucial for seasonal and deep geothermal applications; a comprehensive geological feasibility survey should be ordered immediately to support key planning decisions regarding shallow and deep borehole fields. 3. MIT should study the choice between hot/chilled water versus an ambient loop district system with an open mind. Experience from other university campuses shows that either option will be disruptive. Yet, bold and decisive near-term action is needed to reach our goal. 4. To prepare for the deployment of nuclear batteries or deep geothermal boreholes, MIT must communicate these pathways with Cambridge and state officials and socialize the ideas amongst the broader Cambridge and Boston population from an early stage. In addition, discussion with the Nuclear Regulatory Commission must be started immediately to better assess the project feasibility. 5. To understand whether carbon capture is a valid option for MIT to rely on, potential sequestering application or sites within New England should be explored and future rights secured.
Table of Contents Introduction …………………………………………………………………………… 1 Integrated Model ………………………………………………………………...…... 3 Future Grid Emissions and Climate ………………………………………………… 6 Building Retrofits …………………………………………………………... 9 Advanced Building Controls ………………………………………………. 16 Energy Efficient Labs ……………………………………………………… 20 Geothermal District Systems ………………………………………………. 27 Energy Storage …………………………………………………………….. 36 Nuclear Batteries …………………………………………………………... 43 Deep Geothermal ………………………………………………………… 50 Local Carbon Capture ……………………………………………………... 56 Public Perception and Financial Risk …………………………………….. 63
Members of the Decarbonizing the MIT Campus seminar.
1 Introduction This document is the result of a seminar entitled MIT 4.s42 Decarbonizing the MIT Campus taught during spring 2024 at MIT. The seminar complemented the efforts of an institute-wide working group on the same topic and many members of the working group served as guest speakers in the seminar. Our common goal was to Evaluate what combination of technology upgrades (future and existing) will lead to net zero scope 1 and 2 emissions for MIT’s buildings in what time frame, at what cost and perceived risk. Seminar participants worked with members of the working group to develop comprehensive descriptions of nine technologies – some available today, some not – and two external factors, a local carbon tax and future New England grid emissions, that could influence MIT’s ability to reach its on-site decarbonization goals (Figure 2). Figure 2: Technologies and external factors that influence MIT’s ability to decarbonize its campus The group collaborated on an integrated, physics-based model to evaluate the combined effect of these technologies on campus. We also administered a pilot survey on how the adoption of these technologies might be perceived by the MIT community. The technology descriptions, preliminary survey results and integrated model are described in this document. As shown in Figure 3, over 56,000 decarbonization pathways have been explored. The recommendations presented in the executive summary are based on an analysis of these results and further motivated in the various technology sections. The curious reader is invited explore the decarbonization pathways via an interactive dashboard.
2 Figure 3: 56,000 carbon futures for the MIT campus
3 Integrated Model Nada Tarkhan and Sam Wolk Our model consists of two components, a simulation engine to calculate a large combination of technology pathways for the campus and an interactive dashboard that allows a general audience to browse the pathways and form their own opinion. Engine A simulation engine has been developed with two main steps, building load calculations and energy supply modeling. Building load calculations are executed via the US Department of Energy’s EnergyPlus whole building simulation engine. EnergyPlus is a validated, physics-based simulation to solve heat and mass flow rates in buildings (NREL 2024). The campus model of building loads is a bottom-up portfolio model, developed with urban building energy modelling (UBEM) methods, as in (Reinhart et al. 2016; Ang et al. 2022), and informed by previous work by (Nagpal, Hanson, and Reinhart 2019). EnergyPlus is used to model individual building hourly loads for heating, cooling, hot water, and electricity, informed by building characteristics, and future weather files (see section on Future Grid Emissions and Climate). Each building of interest (buildings owned and operated by MIT), is modeled as a series of 4 representative “shoeboxes,” which serve as a simplification of building geometry for rapid portfolio modeling, while maintaining geometric and solar context. Building parameters are based on calibrated findings from (Nagpal, Hanson, and Reinhart 2019), with individual definitions of building construction. Internal loads are defined based on program, with “templates” defined for circulation/support, residential, classroom/office, and lab space. In the various demand-reduction pathways, envelope retrofits, building controls, and lab retrofits are all modeled as transformations of the underlying EnergyPlus IDF models for each building. Each building is scheduled for renovation according to a sequence which completes all buildings by 2050. Hourly demands for each year are determined by selecting the appropriate result for each building according to its upgrade state, the appropriate year, and climate scenario. By default, all buildings are connected to the combined heat and power (CHP) central utility plant (“the CUP”). In scenarios including development of a new district heating and cooling network, buildings are scheduled for disconnection from the existing CUP network to the new network between 2030 and 2040. The existing CUP network is assumed to have distribution losses of 50% while the new high-temperature district network (“partial” scenario) is assumed to have distribution losses of 10%. Campus electricity demand is first supplied by any carbon-free generation capacity (photovoltaics, nuclear microreactors, and deep geothermal) if available in the specified scenario. Then, a simplified model of the CUP’s equipment (gas turbines, heat recovery steam generators, steamdriven chillers, and electric chillers) is operated at each timestep to meet the total electricity and thermal demands on the CUP, yielding total gas demand with excess electricity imported from the grid when applicable. This model is configured from simple efficiencies and capacities of
4 equipment provided by MIT Facilities and rudimentary control strategies for chiller balancing and HRSG firing. Gas demand is converted to emissions according to a standard emissions factor of 182 kgCO2eq/MWh. The total electricity demand due to the new district heating and cooling networks are computed using constant coefficients of performance for the high-temperature loop scenario and coefficients of performance which vary according to hourly net thermal demand in the ambient loop scenario: heating efficiency improves in the summer while cooling efficiency worsens, and vice versa. In both district demand scenarios, additional energy necessary to balance borefield temperatures is computed according to annual net thermal demand: heat-injection is required in heating dominated years while heat-rejection required in cooling dominated years. Electricity requirements of the district heating and cooling networks are met first by local carbon-free generation capacity if available, and then grid-imported electricity. When applicable, thermal demand is first decreased by any local thermal generation capacity (nuclear micro-reactors or deep geothermal). Grid emissions factors and costs are computed hourly, while gas cost data is computed annually. Dashboard The dashboard visualizes the decarbonization pathways and aims to allow a larger audience to interact with the results. The simulated (over 56,000) decarbonization pathways can be explored through the developed interactive dashboard. This tool is designed to showcase the combined effect of various technologies on the campus emissions and facilitate extracted insight for decision making. The main view presents an aggregate emissions line plot which captures data from over 56,000 scenarios. Utilizing Java and D3.js, this section allows users to interactively explore different emission outcomes over time (from now till 2050) through a combination of preset and customizable scenarios. This is displayed in an interactive line plot. At the top of the dashboard, a dynamic filtering interface, lets users select and combine technologies to see their implementation extent (baseline, partial or full). Users can choose preset scenarios that are hard coded to plot the emission trajectories of the “Business as usual” scenario or “Best practice implementation”. The user can also build their own scenario through recording the generated lines from their explorations. The activated filters influence the display of schematic campus designs in the bottom pane, aligning with the selected technologies. Each layout dynamically adjusts to reflect the chosen scenarios, providing a visual understanding of the potential emissions impact by 2050. Additionally, the filtering mechanisms leverages Java's robust capabilities in handling complex data operations to enable real-time filtering of scenarios based on key metrics such as emissions, cost, risk, and innovation. This part of the dashboard uses D3.js for visual interactivity, enhancing user engagement through brushing and linking techniques that highlight the relationships within the data. The cost, innovation and risk metrics will be integrated once data is available in the future.
5 Development Environment and Tools The dashboard leverages JavaScript and D3.js to manage dynamic data handling and visualization efficiently. It manipulates the Document Object Model (DOM), handles user events, and renders interactive elements, allowing seamless and intuitive interaction with real-time data responses. The interactivity is facilitated through buttons, toggles, and filters, enabling users to reset settings and switch between different scenarios effortlessly. Backend data processing is handled by Java, which manages large datasets and performs real-time manipulations based on frontend inputs. Additionally, D3.js enhances the visual interactivity with techniques like brushing and linking, helping users visually explore and understand data relationships. The dashboard was developed in WebStorm, an integrated development environment (IDE) that specializes in JavaScript, HTML, and CSS, providing robust tools for coding, debugging, and testing. References Y Q Ang, Z M Berzolla and C Reinhart, “From concept to application: A review of use cases in urban building energy modeling, “Applied Energy, 279:1, 2020 S Nagpal, J Hanson and C F Reinhart, “A framework for using calibrated campus-wide building energy models for continuous planning and greenhouse gas emissions reduction tracking,“ Applied Energy, 241, pp. 82-96, 2019 National Renewable Energy Laboratory (NREL), EnergyPlus version 24.1.0, URL: https://energyplus.net/, last accessed May 2024 C F Reinhart and C Cerezo Davila, “Urban Building Energy Modeling – A Review of a Nascent Field,” Building and Environment, 97:196–202, 2016
6 Future Grid Emissions and Climate Ben Taube and Alison Maguire Technology Overview As the generation mix of electricity in the ISONE region and future demands for electricity change, so will the carbon emission rate of electricity. While changes to the regional grid are outside of MIT’s direct control, it is necessary for MIT to consider regional grid emissions in choosing a path toward decarbonization due to the limited options for MIT to generate carbonfree electricity on campus. At the same time, we know that the climate in Boston is going to warm over the coming three decades independent near and long-term actions. This warming will naturally shift our building conditioning loads from heating to cooling dominated. Risk and Innovation Since it is unclear how fast the New England grid will decarbonize, there is a risk for MIT to rely on overly optimistic future emissions numbers. Another risk is that a near fully decarbonized grid may become less stable, introducing additional resiliency concerns. Grid Emission Scenarios To consider a range of plausible future grid emissions scenarios, we picked three hourly marginal emissions rates come from the National Renewable Energy Laboratory’s 2023 Cambium data set (Gagnon et al, 2023) which provides simulated grid emissions in the ISO New England region every 5 years. Cambium also provides hourly electricity prices for each scenario. We used gas prices from 2023 EIA Annual Energy Outlook. The reference case was for business as usual and 95% decarbonization while the low natural gas prices scenario was matched to the same scenario between Cambium and EIA AEO. Table 1: Selected Cambium grid emission scenarios Scenario name Description Business as usual (Mid case) Central estimates for inputs such as technology costs, fuel prices, and demand growth. No nascent technologies. Electric sector policies as they existed in September 2023. Low natural gas prices The same set of base assumptions as the first scenario, but where natural gas prices are assumed to be lower. 95% decarbonization by 2050 The same set of base assumptions as the first scenario, but nascent technologies are included and there is a national electricity sector decarbonization constraint that linearly declines to 5% of 2005 emissions on net by 2050.
7 Local Climate Scenarios The hourly weather data used for the integrated model are based on weather data for Boston. In order to be aligned with the above listed Cambium future grid emissions scenarios, we use the 2012 AMY weather file for Boston Logan International Airport which is the basis for the supply and demand profiles in the Cambium scenarios. This is necessary so that extreme heating and cooling periods coincide between the grid and the campus model. The weather file was then morphed for two emission scenarios using the WeatherShift web app developed by Arup. Morphing is a process where a current hourly weather file is combined with a global climate change models so that hourly temperature, relative humidity and solar radiation can be shifted to new levels for different future years. Table 2: Future climate file scenarios Scenario name Description Emission scenario RCP 4.5 (moderate) An intermediate scenario in which emissions peak in 2040, then decline. Most probable baseline scenario without climate policies. Emission scenario RCP 8.5 (extreme) An aggressive scenario in which emissions continue to rise through the end of the century. Thought of as a worst-case scenario. References M. Bamm, et al (2017) WeatherShift Water Tools: Risk-based Resiliency Planning for Drainage Infrastructure Design and Rainfall Harvesting SE Belcher, et al (2005). Building Services Engineering Research and Technology, Constructing Design Weather Data for Future Climates Gagnon, Pieter, Pedro Andres Sanchez Perez, Kodi Obika, Marty Schwarz, James Morris, Jianli Gu, and Jordan Eisenman. 2023. Cambium 2023 Scenario Descriptions and Documentation. National Renewable Energy Laboratory U.S. Energy Information Administration. 2023. Annual Energy Outlook AEO2023
8 Vacuum Insulation Panel in MIT’s recently renovated Building 2 (Image: E Reinhard)
9 Building Retrofits Zoe Le Hong, Yewon Ji, and Ray Wang Technology Overview Building retrofits represent a strategic approach to enhancing the sustainability and energy efficiency of existing structures. This process involves applying various technologies and practices to improve overall performance. By implementing measures such as optimizing insulation, enhancing airtightness, upgrading HVAC, and integrating renewable energy sources, building retrofits can substantially mitigate energy consumption and reduce carbon emissions. Retrofitting will reduce buildings' heating, cooling, and electricity loads, thus reducing energy demand (and associated costs and emissions) and stress on MIT’s district energy and electricity systems. Additionally, retrofit projects can enhance occupant comfort and indoor air quality, contributing to a healthier and more productive campus environment. We systematically assess MIT's existing building portfolio. This involves gathering pertinent data and creating energy models to establish baseline building conditions. Subsequently, we categorize buildings based on age and program type, following the current groupings of the MIT Facilities teams. These groups inform model inputs of material characteristics, programmatic function, system operations. Using this information, we propose potential typology-level upgrades aimed at progressively reducing buildings' energy loads over time. For each building, we will establish the baseline model as reference, calibrated against actual metered data. For each upgrade scenario, we consider the following KPIs: - Identification of potential retrofit measures with consideration of technical compatibility, risk, budget, levels of disruption, and lifecycle emissions - Combining these measures into packages of 2 upgrade scenarios (partial and full implementation) for different retrofit options (ex. façade, windows, etc.) - Baseline & scenario average Energy Use Intensity (EUI) by end-use – heating, cooling, electricity (KWh/m2), with % change in EUI Risk and Innovation The Capital Renewal Team at MIT has historically conducted building assessments every 5 to 6 years and devised comprehensive programs every 10 years for the entire campus. However, these initiatives have not primarily focused on energy usage, and given the limited space on campus, the priority for most buildings is to accommodate laboratories, classrooms, offices, and similar facilities. The campus is particularly sensitive to disruptions, especially in residential halls and lab-intensive buildings. Implementing these changes risks disrupting the allocation of space for current programs, a concern that poses substantial difficulties for full-scale implementation measures. Even partial implementation efforts must be carefully planned, often necessitating that work be conducted during the summer months when the campus is less populated and thus less disrupted by such activities.
10 Current State of Buildings To better understand opportunities for introducing efficiencies and viable energy reduction measures, as well as develop a calibrated baseline energy model, the current energy use of (metered) buildings on campus were reviewed. Metered energy use data from the MIT Sustainability Data Pool at the building scale from 2010 were analyzed. Some key findings are summarized: - Labs are estimated to use about 35% of total campus building use - LEED buildings have marginally lower average normalized energy use (see Figure 4). - On average, across buildings which underwent capital improvements, gas and steam consumption significantly decreased, but cooling and electricity use increased. - Several buildings which underwent significant capital improvements and/or energy retrofits show little change in energy use after renovations were complete (for example, Building 9 and Building 2, seen in Figure 5 and 6). o However, for Building 9, in the event of energy retrofits, while total energy use does not change much, peak loads decrease. o Additionally, for Building 2, an increase in electricity use after renovations could be due to higher occupancy rates after upgrades occur. - Tang Hall (Building W84), which underwent a thermal control upgrade in 2021, exhibits a clear decrease in heating loads. - Buildings with higher total energy use are not necessarily buildings with the highest energy use intensity (per square foot). This can be an important metric to consider when targeting buildings for retrofits which can achieve the most emission reduction. In addition to understanding current energy use, through conversations with the MIT Facilities teams, we gained insight into current building operations and capital improvement functions across MIT-owned buildings. While this analysis was not comprehensive, some critical findings contributed to our recommendations to expand Facilities teams’ capacity to enable campus retrofits at scale. In particular, shifting capital investments from reactionary to anticipatory (or preventative) through more frequent data collection and analysis is seen as an important factor that will enable improved operations and more efficient retrofit diffusion across campus buildings. This is discussed further in recommendations.
11 Figure 4: Histogram of LEED and Non-LEED building energy use intensity (total energy use normalized by building area) Figure 5: Annual energy use of Building 2, by end-use. Figure 6: Annual energy use of Building 9, by end-use.
12 Scenarios Scenario name Description Cost [$] Space requirement [m2] Risk [credits] Innovation [credits] Business as usual A model representing current building energy use and load demands for all buildings owned and operated by MIT on the main campus. Energy Star Appliances Academic/ A rating / Cost Not Available / Maintenance cost Not available Cool roof system (reflective film) / 10% energy None 1/8” 0 0 Partial implementation [1] Installation of nonintrusive retrofits, allowing building operations to continue throughout capital upgrades with minimal to no interruptions. Energy Star Appliances Residential / A rating Reflective window adhesive films / $14 per sqft PV / Inverter upgrades / 90-95% efficiency Air source heat pump 3 – 5 COP / $112 per sqft Active Chilled Beams / 30-40% savings / $47 per sqft LED Lighting Academic/ 75% efficiency / $15 per sqft LED Lighting Residential / 75% efficiency / $15 per sqft Low-flow plumbing / 2GPM / $1900 per equip. Duct Replacement & cleaning / $42 per sqft None None 2 sqm 2 sqm Up to 12” on ceiling None None None None 1 0 Full implementation Full energy retrofit, with no limitations with regards to operational disruptions (for example, installation of additional insulation in walls). Wall insulation upgrade / R-20 / $150 per façade sqft Curtainwall upgrade / U-0.2/ $79 per facade sqft Window Upgrade / U-0.2 / $79 per façade sqft Air-barrier Upgrade / <0.1perm / $40 per façade sqft Roof insulation upgrade / R-40 / $31 per roof sqft Radiant System / 6 to 11 W/m2/ $84 per sqft Heat Recovery Systems / 3 COP Low-flow plumbing / 2GPM / $1900 per equip. New PV installation / 5kWh/m2 / $61 per roof sqft 4”, disruptive 6”, disruptive None, disruptive 1”, disruptive Additional 4” 5” floor depth 9m3 per unit None Available roof area 2 0 [1] The “sensitivity” to disruption and energy needs are unique to each building archetype, so retrofit packages differ between archetypes. However, as the baseline model is somewhat calibrated, the change in energy use for each scenario is different for each building.
13 In the field of retrofitting technology, innovative solutions are scarce because most retrofit technologies, like facade upgrades, HVAC improvements, or transitioning to LED lighting, are already widely available. Model The building energy model described in the integrated model, is used to evaluate the campus energy demands for the three given decarbonization scenarios. For these representations, buildings are divided into 12 representative archetypes, by dominant use-type (as is currently defined by MIT Facilities of lab-, classroom/office -, residential-, and support-dominant), and 3 age groups (pre-1945, 1945-1980, 1980-2015). In each scenario, retrofit “packages” are developed for each archetype with an estimated cost and timeline normalized by floor, façade, window, or roof area. Buildings are assumed to be renovated at a rate of 4% (necessary to retrofit all buildings by 2050), with buildings selected based on a “priority” score (related to age and past campus data). If a building is selected for a retrofit, the retrofit packages are implemented all at once (no phasing). As described above, scenarios are based on allocating different scales of retrofits (rather than an alteration of retrofit rate), and are implemented by dynamically editing the campus model EnergyPlus IDF files for groups of building archetypes. Key Recommendations To reduce building emissions in a significant way and make the most out of the (unavoidable) high financial burden of conducting retrofits at scale, we suggest that lab buildings are focused on (see Sec. Energy Efficient Labs), and that, while more disruptive, deep retrofits are far more worthwhile to invest in. In the interest of reducing demand in a sustainable and financially prudent way, we propose that MIT focus not on the low hanging fruit of shallow retrofits first, but in building the capacity within the university to (1) better understand the current status of building energy use on campus, and (2) maximize efficiencies in the retrofit process through comprehensive planning. Thus, we suggest the implementation of a Buildings Energy Team through an expansion of the existing MIT Facilities teams. For example, the current teams do not have the resources to consider energy usage before and after building upgrades and have not been able to quantify the benefits that come from retrofits. We think that if more focus is put on these benefits (ecological, monetary, health, etc.), “disruptive” interventions may be more accepted. This team could contribute to identifying and planning for efficient retrofit installations at scale and identify operational efficiencies that may not require physical energy conservation measures. While we have identified that building energy retrofits are not perceived as innovative, accomplishing a large-scale retrofit plan in an efficient manner with special consideration for operations and disruption would be truly market leading. Finally, the team could spearhead building emissions reduction education and outreach across MIT, which we believe is critical to ensuring MIT’s building retrofit investments are both accepted and celebrated internally and externally.
14 Cost - Cost of partial: $849,447,811 (~ $6.6M per building, $502 per m2 of TFA / $47 per ft2) - Cost of full: $2,577,299,872 (~ $26.3M per building, $1,994 per m2 of TFA / $186 per ft2) The estimated cost includes construction expenses such as removal of existing systems, staging, remediation costs, labor, construction management costs and design contingencies. Costs from disruptions (such as downtime, relocation, and vacancy), and soft costs are not accounted for. Additionally, costs associated with not retrofitting are imminent and relevant with approaching local regulations (BEUDO). Due to tax parcel deliminations, most buildings on MIT campus will be subject to fines in the coming years, which can increase costs further. This could incentivize a more rapid retrofit rate across campus buildings. [2] Building systems are not modelled explicitly. Heating and cooling loads assume a coefficient of performance of 1. Further Reading be-exchange.org/beexreport/commercialdata/ https://static1.squarespace.com/static/5d375710088ee600019ba2d9/t/64c2aeb85f8cbc7a6c993 69b/1690480313209/CLRC+Union+Labor+Costs+in+Construction-2022.pdf References Ang, Yu Qian, Zachary Michael Berzolla, Samuel Letellier-Duchesne, Violetta Jusiega, and Christoph Reinhart. 2022. “UBEM.Io: A Web-Based Framework to Rapidly Generate Urban Building Energy Models for Carbon Reduction Technology Pathways.” Sustainable Cities and Society 77 (February): 103534. https://doi.org/10.1016/j.scs.2021.103534. Crawley, Drury, Frederick Winkelmann, Curtis Pedersen, and Linda Lawrie. 2011. “EnergyPlus: A New-Generation Building Energy Simulation Program.” In , 58–59. US Department of Energy. https://dl.acm.org/doi/10.1145/1961678.1961690. Nagpal, Shreshth, Jared Hanson, and Christoph Reinhart. 2019. “A Framework for Using Calibrated Campus-Wide Building Energy Models for Continuous Planning and Greenhouse Gas Emissions Reduction Tracking.” Applied Energy 241 (May): 82–97. https://doi.org/10.1016/j.apenergy.2019.03.010. Reinhart, Christoph F, Timur Dogan, Alstan Jakubiec, Tarek Rakha, and Andrew Sang. 2013. “UMI - An Urban Simulation Environment for Building Energy Use, Daylighting and Walkability.” In BS2013. Vol. 13. Chambéry, France.
15
16 Advanced Building Controls Akshata Atre and Ben Taube Technology Overview Distributed digital sensors and controls for temperature, humidity, CO2, occupancy, and other factors can reduce energy use in buildings by calibrating the heating, cooling, and ventilation of spaces to match measured occupant levels. Instead of heating and cooling entire buildings or building stories to predetermined setpoints, these controls enable us to turn mechanical systems on or off on a more detailed zone-by-zone, or even room-by-room, basis, depending on actual use patterns. Refining scheduled setpoint and setback temperatures can provide significant energy savings and improve thermal comfort while ensuring that spaces are being conditioned properly (Papadopoulos, 2019) alongside energy savings from providing fresh air only when it is necessary. Technology to implement advanced building controls is already commercially available in both off-the-shelf and customizable configurations. The main challenges to deploy such controls across MIT’s campus are careful, long-term planning to minimize disruption to institutional activities during the installation process as well as data security concerns. Fortunately, given that these types of controls and systems have been in wide use for a long time, there is no need for MIT to conduct preparatory research on whether they work and/or are effective. To fully benefit from this technology and reduce installation costs, better coordination is required to ensure advanced controls are part of all building retrofit projects, reducing installation costs. In residential construction, modern controls typically pay for themselves in less than one year. These numbers do not translate directly to commercial buildings since building management system (BMS) enabled sensors can be an order of magnitude more expensive than residential smart thermostats. However, switching to wireless controls can save on costs significantly, and once advanced controls are in place, there are significant opportunities for MIT researchers to innovate. In fact, MIT is currently conducting a study on AI-based controls in select classrooms which shows that innovative control strategies increase savings without any additional hardware costs (MIT News). Evolving sensing technologies such as those currently being developed at the Media Lab (“Responsive Environments Group”) could also reach their full potential if tested at the campus level, contributing to fundamental research while saving on-site carbon. Advanced digital controls can also be used for lighting and other building energy end uses to further reduce campus carbon emissions. Risk and Innovation High resolution temperature controls are a very low-risk technology that will greatly benefit the campus’s decarbonization efforts. The technology is widely available, and the potential cost and carbon savings are worth the effort of implementation. Indeed, MIT faces a reputational risk if it continues utilizing existing mechanical controls, which are outdated and energy-intensive while other peer institutions such as Stanford (de Chalendar, 2023, Hu, 2023) take steps to modernize and optimize HVAC. MIT will also likely incur undue operational costs, unnecessarily conditioning unoccupied and unused buildings as seen in 2020 during the COVID-19 pandemic when energy
17 use for heating and cooling our building remained constant during the spring of 2020 even though out campus was largely deserted. Scenarios Scenario name Description Cost [$million] Risk [credits] Innovation [credits] Business as usual Implementation of advanced temperature controls in existing buildings is largely ad hoc and inconsistent. Varies, assume zero for model 0 0 Partial implementation By 2030, implementation of scheduled temperature setbacks and outside air supply rate reduction calibrated to schedulebased occupancy of spaces on campus, and potentially during a holiday curtailment break. These changes will coincide with buildings scheduled to be retrofitted within the same time frame. $1,500/ther mostat 0 0 Full implementation By 2040, implementation of scheduled temperature setbacks and outside air supply rate reduction in all buildings on campus, in addition to the deployment of sensing technology or occupancy-based controls for more highly calibrated and responsive heating and cooling. $1,500/ther mostat 0 1 Model To model the effect of scheduled thermostat controls that are more aligned with building occupancy, we assign a date of retrofit to each building on campus as noted in the scenario descriptions. Each upgraded campus building gets a new control schedule instead of the default constant setpoint thermostats in older buildings and uses occupancy-based ventilation controls to reduce outdoor air supply when buildings are unoccupied. Because our building energy models do not have detailed zone information or HVAC systems, higher spatial resolution controls and more advanced controls such as temperature resets. The figure below shows typical weekday schedules for existing and renovated buildings.
18 Figure 7: Typical weekday schedules for existing and renovated buildings. Key Recommendations MIT does not currently support wireless temperature controls in buildings. Going forward, Facilities should establish wireless sensing and control protocols to meet the security needs of MIT and a central retrofit coordination group should ensure that controls are installed in all renovation and retrofitting projects with a plan for full campus implementation by 2040. This task would fall under the purview of the Retrofit Central team suggested in the Executive Summary. References de Chalendar, J. A., Keskar, A., Johnson, J. X., & Mathieu, J. L. (2023). Living Laboratories can and should play a greater role to unlock flexibility in United States commercial buildings. Joule, 8(1), 13–28. https://doi.org/10.1016/j.joule.2023.11.009 Hu, M., Rajagopal, R., & de Chalendar, J. A. (2023). Empirical exploration of zone-by-zone energy flexibility: A non-intrusive load disaggregation approach for commercial buildings. Energy and Buildings, 296. https://doi.org/10.1016/j.enbuild.2023.113339 MIT News | Massachusetts Institute of Technology. “AI Pilot Programs Look to Reduce Energy Use and Emissions on MIT Campus,” September 8, 2023. https://news.mit.edu/2023/ai-pilotprograms-look-reduce-energy-use-emissions-mit-campus-0908. Papadopoulos, Sokratis, Constantine E. Kontokosta, Alex Vlachokostas, and Elie Azar. “Rethinking HVAC Temperature Setpoints in Commercial Buildings: The Potential for Zero-Cost Energy Savings and Comfort Improvement in Different Climates.” Building and Environment 155 (May 2019): 350–59. https://doi.org/10.1016/j.buildenv.2019.03.062. “Responsive Environments Group - MIT Media Lab.” Accessed April 30, 2024. https://resenv.media.mit.edu/.
19
20 Energy Efficient Labs Amanda Kirkeby Technology Overview Laboratories play a vital role on every research campus, providing a space for researchers to work towards solutions for some of the world’s most challenging issues. At MIT, with a global reputation for challenging the status quo, defying limits of technology, and pushing the very boundaries of human knowledge, cutting-edge research occurs every day in the 180,000 square meters of laboratory space on campus. Especially in laboratories where the research is environmentally sensitive or potentially hazardous to human health, the building systems in laboratories are the primary protection mechanism in mitigating risk to the most important asset of the MIT campus: its world class researchers. With the crucial role building systems serve in laboratories, labs have an energy use intensity much higher than other academic, ancillary, or residential space. In a MIT campus modeling effort, Nagpal et al. estimated that special use laboratory spaces have an energy use intensity 10 times that of office spaces (Nagpal, Hanson, and Reinhart 2019). Main drivers of this significant amount of energy use are ventilation and equipment. With air being the primary defense from research-based contaminants, ventilation-driven loads typically account for upwards of 65% of total laboratory energy use (Kirkeby 2022). Research equipment loads can account for 20% of total energy use (Kirkeby 2022). As a result, laboratory-dominated buildings are the leading contributors to campus energy consumption, making energy efficiency improvements in laboratories a key component in decarbonization efforts. However, due to the “mission-critical” nature of lab operations, it is vital that all strategies proposed must: - Prioritize safety of researchers and the broader MIT community, - minimize disruption to research and - provide ongoing carbon savings through a programmatic approach. While no small undertaking, there are ample strategies for optimizing the safety and efficiency of research laboratories. For the past 20 years, efforts by the U.S. Department of Energy (DOE) Federal Energy Management Program (FEMP), the International Institute for Sustainable Laboratories (I2SL), and leading research institutions have been developing strategies to improve both the energy efficiency and the effectiveness of laboratory building systems to provide a safe working environment for researchers. A leader in this field, the University of California – Irvine (UCI) Smart Labs™ program was recognized by the DOE and NREL as the flagship program for Smart Labs, showing it is possible to successfully address the significant undertaking of efficient laboratories in high-profile research institutions (NREL and I2SL 2024). This spurred the Better Building Smart Labs Accelerator, hosted by DOE FEMP and researchers at NREL in 2016. The Smart Labs Accelerator challenged 17 U.S. research laboratories to reduce energy use by 20% (U.S. DOE Better Buildings 2020). With the success of the Smart Labs Accelerator, the Smart Labs Toolkit was developed by NREL to facilitate decarbonization efforts by research institutions across the globe, curating lessons learned by accelerator participants and best practices from leaders in laboratory sustainability, including I2SL, 3Flow, and My Green Lab (NREL and I2SL
21 2024). The Smart Labs Toolkit presents a programmatic approach that involves key stakeholders from management to facilities staff to environmental health and safety personnel to researchers. Through a systematic four stage process – Plan, Assess, Optimize, and Manage – key facilities are assessed to identify energy conservation measures that address key contributors in energy consumption: ventilation and equipment. Establishing a Smart Labs program rather than completing a one-time efficiency improvement effort supports the dynamic process of assessing, optimizing, and managing required to maintain efficient, safe operations in such a dynamic environment as laboratories. Through discussions with the MIT Facilities and Environmental Health & Safety (EHS) staff, the current efforts towards energy efficient laboratories in development are already in alignment with many of the best practices set forth by leaders of the sustainable laboratory community. These improvement measures have just been implemented in Building 76, the leading building in energy consumption on campus, and plans exist for similar projects across the rest of campus. In this section, we propose three implementation scenarios for addressing energy efficiency and carbon savings in laboratories. Brief descriptions of the key strategies are included below, with further information on cost, risk, and innovation in the scenarios table in this section. These considerations are laboratory-specific energy conservation measures found through the detailed audit performed in Building 76 and other considerations from Smart Labs best practices. Summary of Laboratory-Specific Efficiency Strategies Optimized Airflow Based on Ventilation Risk Assessment - Reduce ventilation to flows appropriate for research-specific hazards through a Ventilation Risk Assessment (LVRA), which considers specific hazards posed by research in each space to assign tailored, appropriate ventilation rates that effectively mitigate risk of contaminant exposure to researchers. - Repair laboratory ventilation and airflow setback controls and address deferred maintenance to get system operation back to effectiveness as originally designed. - Implement demand-driven controls, allowing the right airflow to be provided in the right place at the right time. This will greatly reduce unnecessarily high energy consumption during unoccupied periods. Informed Selection and Placement of Supply Diffusers - Relocate supply diffusers and general exhaust grilles to promote ventilation effectiveness and the sweeping of contaminants away from researchers. High Performance Exposure Control - Replace constant volume fume hoods with high performance variable air volume (VAV) models when possible, to allow for flows to modulate appropriately with use. - Implement program or engineering controls to shut fume hood sashes when not in use to prevent unnecessary high flow rates when no hazards are present. Energy-Efficient Equipment - Adjust setpoints for Ultra-Low Temperature Freezers, which can save significant amounts of energy consumption without affecting stored samples. - Develop a standard of minimum efficiency requirements for new equipment procurement with researchers to reduce research equipment energy and water consumption.
22 - Implement a space and equipment sharing program, which can minimize future needs for additional laboratory space. Perhaps the most important thing to note across all these strategies is to engage researchers in the process. This gives them agency in reducing the impact of their own research, rather than imposing seemingly restrictive regulations on their work, which has been found to have a much higher success rate (My Green Lab 2024). It also enhances the reputation of the conditions for researchers at MIT, enticing new top talent to seek out research opportunities at this esteemed institute. With all this in mind, such efforts require significant coordination across all stakeholder groups within MIT, from researchers to EH&S and Facilities staff to management. Beyond the technological strategies posed, it is highly recommended that dedicated staff that are knowledgeable on MIT processes and sustainable laboratory principles are assigned to coordinate these efforts. Risk and Innovation The implementation of these strategies poses very little risk to MIT. Some of them may require more financial investment than others, but the tried-and-true nature of many of these strategies in the sustainable laboratory global community make them low-risk to implement. On the contrary, not implementing the proposed strategies pose a significant risk to the Institute on multiple levels. Perhaps most importantly, the ineffective operation of laboratory ventilation can pose a health threat to MIT’s most important asset: its researchers. Across the country, 1 in 3 laboratories pose a significant risk to the health and safety of its researchers due to ineffective airflow (3Flow and I2SL 2016). This statistic even applies to cases where ventilation systems are properly designed, since deferred maintenance can lead to 50% reduced effectiveness in as little as 5 years (NREL 2020). Additionally, the added cost of potentially wasted energy consumption from this ineffective operation can be as much as $5 more per square foot to operate than officetype spaces (Better Buildings Alliance - Laboratories Project Team n.d.). By implementing these proposed strategies, not only will it enhance the safety of the working environment for researchers; the significant operational cost savings can be devoted to further improvements of facilities. Furthermore, the reputation of MIT as a research institution is at stake in not implementing these strategies. The Institute is falling behind many peer research universities who have already implemented sustainable laboratory best practices and strategies that prioritize both safety and efficiency. As a global research institution, the delay of implementing these strategies not only poses the risk of not achieving the status of a “zero-carbon” campus; there is the potential for this to reduce to the influx of potentially top-tier researchers who are deterred by facilities that are operating sub-optimally. In short, implementation of these proposed laboratory efficiency improvements and building system optimization are needed to maintain a safe working environment for current researchers and enhance the Institute’s reputation as a global leader in cutting-edge research and innovation.
23 Scenarios Scenario name Description Cost [$] Space requirement [m2] Risk [credits] Innovation [credits] Business as usual No changes to ventilation or any systems. Risk to researchers due to systems not operating as effectively as designed. 0 NA 3 0 Partial implementation* Optimize airflow, minimal disruption to research. Reduce ventilation rates to appropriate levels based on risk assessment (LVRA) and repair airflow programming, including implementing occupancy controls. $5,000,000 $27/m2 NA 1 2 Full implementation* Optimize airflow and address equipment efficiency – potential for short-term disruption to research. Reduce ventilation based on LVRA and fix airflow programming, including implementing occupancy controls. Convert CV to VAV fume hoods and implement proper sash management. Relocate supply and exhaust grilles for optimal airflow. Install a Konvecta heat recovery system. Promote an efficient equipment procurement program, including water-efficient autoclaves to reduce hot water demand. $31,400,000 $174/m2 NA (Potential to reduce future space requirements) 1 3 *We assume MIT’s 5,500 labs will be steadily updated until 2050 at a rate of about ~220 labs per year.
web.mit.eduRkJQdWJsaXNoZXIy MjA2MzQ5MA==