Case study of low-cost energy reductions in shared research laboratories at a U.S. public university

Abstract

Scientific sample storage and assay development require vast energy; decreasing the laboratory’s footprint addresses urgent climate needs and efficient use of indirect funds. However, external sustainability certifications require financing and administrative buy-in. We describe our near zero-cost effort to improve energy efficiency within a U.S. public university system.

Methods:

We collected forty-four plug-load estimates of individual machines (e.g., cold-boxes, incubators, hot-plates) using commercially available 120V energy monitors. Measurements occurred for an average of 135 minutes. Larger device data (biosafety cabinets and autoclave) were interpolated from the literature, while ultralow temperature freezer values were estimated based on a combination of measured, and previously reported, consumption estimates. Estimated savings from fourteen discrete interventions are cataloged.

Results:

Our group’s year-long adjustments, particularly estimates from turning up ultralow temperature freezers and powering down biosafety cabinets and an autoclave after hours, resulted in savings similar to powering 9 homes. Our estimate of these savings exceeds 83,000 kWh or ~$10,000. Extrapolation of scenario estimates to campus-wide savings from the freezer intervention suggest University savings of more than 2 years’ of indirect funding costs for an R01 equivalent grant, while annual energy savings for the 430 campus-wide labs may exceed $4 million.

Introduction

Climate change profoundly affects human health (1). As global temperature increases, so does mortality due to a variety of causes, including food insecurity, wildfire smoke, air pollution, and heat. This is especially the case in vulnerable populations, such as the very young or elderly (2).

Morbidity associated with climate change is also increasing. For example, young people are experiencing unprecedented lifetime exposure to heatwaves (3,4). Climate change also affects their mental health. Seventy-five percent of those aged 16–25 perceive the future as frightening and nearly half report their daily life is negatively affected by feelings about climate change; many feel governments have inadequately responded (5). Climate action may lead to feelings of empowerment and alleviate climate distress. For example, community-based interventions including mentoring and environmental interventions are suggested to positively impact mental health (6).

Because the youth are particularly affected by climate change, and are also often involved in university settings, we engaged young scientists in climate action at a research university. Research laboratories are very energy intensive, requiring roughly 10 times the energy consumption of a typical office building (7,8). Work from the Environmental Protection Agency in 2008 estimated that if only half of U.S. research laboratories improved their efficiency by 30%, the equivalent energy consumption of one million households would be saved (9). This is due to the large energy consumption of common lab equipment. For example, one biosafety cabinet (BSC) consumes roughly half the energy of a U.S. household, ultralow temperature (ULT) freezer one household-worth, and autoclave sterilization over three households’ (10–13). Laboratories also generate millions of tonnes of plastic waste (14). Other research groups have identified the urgency of reducing the carbon footprint of academia (15) and provide in-depth efforts made in European and the United Kingdom research institutes (16–18), but literature on voluntary initiatives in the U.S. is lacking.

Institutions should be motivated to improve efficiencies in research for at least two reasons – these policies save significant indirect costs and funders are increasingly requiring proof of environmental stewardship in research practices (19,20). Herein, we report our approach to significantly decrease lab energy use without financial buy-in at the institutional level.

Methods

Our Lung Biology and Cystic Fibrosis Research Center focuses on a wide variety of adult and pediatric pulmonary disease processing using wet- and dry-research lab techniques. The open-floor plan laboratories are occupied by 19 PIs and roughly 100 staff, faculty and students.

Communication, group formation

In Fall, 2023, one PI introduced the concept of laboratory sustainability during a floor-wide meeting with assistance from a post-doctoral scholar. Seven months later (2024), an “Earth Day” follow-up meeting was presented by the same PI. A group of interested investigators remained after this meeting and discussed their experiences with and interest in sustainability, forming the University of Iowa Biomedical Sustainability Initiative (BSI) (21), comprised of 11 people from eight labs. BSI meets weekly to discuss targetable solutions.

Energy monitoring

Individual machines in the laboratory were monitored using commercially available 120V energy monitors (Suraielec Watt plug-in meter), generously provided by the University of Iowa Pappajohn Biomedical Institute. We collected forty-four plug-load estimates of common lab equipment (e.g., cold-boxes, incubators, hot-plates) for an average of 135 minutes/device. For device plugs incompatible with 120V monitors, data were interpolated based on measurements of similar sized devices performing similar functions, and publicly available data. To account for ULT freezer heterogeneity, we ascribed savings based on reported literature and direct measurement of one ULT freezer at two setpoints, see details in Supplemental Methods and Table S2; −80°C unit consumption was reported as 14 kWh/day, and −70°C as 10.6 kWh/day. Biosafety cabinets and autoclave values were interpolated from the literature. Three older, large BSCs were estimated to consume 18.8 kWh/day, and a smaller BSC 11.1 kWh/day (10,22). Estimated savings from fourteen discrete interventions are presented.

Electrical costs were determined as 12.05 cents/kWh, based on Iowa’s reported residential rate (January, 2025) (23), and calculated from kWh reported by the monitors using the following formula:

$$\mathit{\text{Annual energy}}\text{c}ost(\$)=\frac{\mathit{\text{Total energy monitored}}\left(kWh\right)}{\mathit{\text{Time monitored}}\left(h\right)}\times 24\frac{h}{day}\times 365\frac{\mathit{\text{days}}}{yr}\times \frac{\$0.1205}{kWh}$$

Annual savings were compared to 9,567 kWh, the average electrical consumption of Midwest homes (11). Findings were reported in a shared Microsoft Excel database, available to lab members.

Interventions

Cold-box cleanout:

Lab managers and PIs were contacted to schedule a cold-box cleanout day to encourage efficient space and resource use. Lab members were trained in responsible waste-stream management prior to disposal of old, unlabeled, and/or expired samples and reagents from shared spaces. Lunch was provided to incentivize participation.

ULT freezer campaign:

Individual PIs were approached and informed of the climate benefits (~35% reduced energy consumption) of increasing ULT temperature settings from −80°C to −70°C, which until ~2010 was standard practice. Additional benefits include reduced machine downtime, reduced utility expenses, and longer freezer lifetime. Public and private universities and industry partners who joined this effort, (according to My Green Lab^® Impact Laboratories) were mentioned. Scale-up projections were estimated from a 2023 University of Iowa asset report of ULT inventory (N=702).

Education to sustain the culture:

New employees were identified, and biannual onboarding meetings scheduled. Meetings consist of a standardized presentation, followed by lab walk-through to provide employees with an overview of energy-saving practices in the lab, particularly BSC, autoclave, and cold-box use. Undergraduate employees responsible for autoclaving were encouraged to combine loads to maximize efficiency.

Visual communication:

Laboratory machines were audited and designated into three groups, those that: 1) should be powered down if not in use, 2) powered down at the end of the day, or 3) never powered down. Stickers were then applied indicating designated category. Flyers were designed to encourage PIs to turn up ULT freezers to −70°C and minimize time spent with freezer doors open.

Statistical analysis:

Directly measured energy consumption of devices drawing power for 24 hours was compared to that of devices drawing power for less than 24 hours by unpaired t test. Analysis performed in GraphPad Prism version 10.4.2.

Results and Discussion

Communication, group formation

Large-group lab meeting presentations served as catalysts for change by facilitating conversations around how to improve efficiency on the floor, allowing various voices and perspectives to be heard. Because the study was performed in a large, shared, open-floor research setting, individuals had previously employed various approaches to ensuring devices were ready and functional at the time of their assays. For example, before our study, some investigators had posted signs on equipment stating: “Never turn off this machine”. This was due to perceived inconvenience (machines take too long to heat-up), and anecdotal equipment malfunctions after power-downs. Therefore, large-group discussion and collective brainstorming were vital to address these concerns and smooth workflows for common laboratory equipment.

These meeting presentations resulted in the formation of a small group of sustainability advocates, which was enhanced by the presence of investigators of all ages and career stages, representing eight unique labs and diverse perspectives. The BSI met weekly to design tangible solutions to improve energy efficiency without impeding data generation and project completion. Topics and results are discussed in a weekly newsletter, see Fig. S1. For a timeline of actions and outcomes, see Table S1. Images from these efforts are found in Fig. S2. Our group observed dynamics that were not measured: 1) Direct in-person small group and one-on-one communication (and repetition) appeared more effective at changing habits than mass-communicative methods, like email; 2) Devices that had been powered-down (primarily by BSI members) remained off for long periods of time. Motivating non-BSI lab members to power-down devices was more challenging; 3) Adoption of large energy saving interventions, such as BSC power-down and −70°C setting was undertaken by more employees. This may be due to a variety of reasons including appreciation of the outsized impact of these actions, because these devices are loud, devices are frequently used (e.g., encountering a powered-down BSC signals to the next user it is OK to turn off after their assay), because the shift to −70°C only occurs once, or for other reasons.

Energy monitoring

Overall, 44 measurements of common laboratory equipment were carried out (Fig. 1a). Each device was measured continuously for an average of six days using commercially available 120V monitors.

Fig. 1. Energy consumption of common lab equipment.

a. BSI member (LT) demonstrates energy consumption assessment of common lab equipment (in this case, a water bath); b. Daily energy consumption and its relationship to the amount of time an item drew electrical current. Directly measured daily energy consumption was variable for devices consuming energy for 24 hours, and did not differ from those consuming energy less than 24 hours (unpaired t test, P=0.39); c. Directly measured daily energy consumption by device category. Heat blocks reported a large range in energy consumption, with both temperature and model affecting annual energy consumption; b-c: Data points representing the same device type at the same temperature set-point indicate different models.

We determined the energy consumption of common laboratory equipment designed to heat samples by analyzing nine machines, Fig. 1b–c. Daily energy consumption, based on time spent drawing current, is shown in Fig. 1b. Because the hot plate was transiently on, its annual consumption was near-zero. Surprisingly, devices powered on for fewer than 12h did not require less energy than those powered on for 24h (0.44 vs. 0.81 kWh, p=0.39), warranting further analyses, Fig. 1c. While heat block temperature influenced energy consumption, the 56°C data demonstrate some models are more efficient than others (for specific models and settings, see Fig. S3).

Heat block findings highlighted the need to turn off these devices after-hours, yet concerns were raised about delays in arriving at setpoints during critical experimentation, resulting in purchase of outlet timers for overnight and weekends. Heat block power-on time was thereby reduced by 79%.

Based on these data, reducing power-on time and selecting efficient and/or sustainable models (which can be found using the My Green Lab^® ACT^® Ecolabel (24)) are important steps to improve energy consumption. We believe manufacturers should report energy consumption of common devices at specific temperatures using a standardized protocol. Switch-timers may be used for non-peak use hours, while for devices potentially requiring constant power (e.g., incubators), efficient use of space is critical to ensure additional devices are not purchased or required.

We estimated total annual savings from our interventions to be approximately 83,000 kWh ($10,000), Table 1. Based on the U.S. Energy Information Administration (11), total kWh savings from these interventions are roughly equivalent to powering nine average Midwest homes. This estimate is based on both measured kWh savings from interventions, as well as interpolated values from the literature (10,12,22), denoted in italics in Table 1. Due to the paucity of available data, we are unable to extract the precise savings experienced by the university, however, if our efforts were expanded to the roughly 430 labs on campus, we have estimated that the university would save roughly $4,300,000.

Table 1.

Measured and interpolated projected energy and financial savings from interventions on a per annum basis. Interpolated values are denoted by italicized font, see details in the footnote section. Financial savings calculated based on a rate of $0.1205/kWh.

Intervention	Averted energy (kWh)	Savings ($)
35 freezers from turned up from −80°C to −70°C (or −75°C)a	~40500	~$4900
Turned off front-loading, steam jacketed autoclave during nights and weekendsb	~23400	~$2800
Shutting sash, turning off 2 biosafety cabinets nights and weekendsc	~8300	~$1000
Surplussed 3 LRP sliding-door refrigerators, replaced with 2 previously purchasedd	4315	$520
Powered down Excella E25 New Brunswick Scientific chest shaking incubator	1026	$124
Surplus Sears Roebuck & Co Kenmore fridge/freezer, replace with American Biotech Supply 2-door refrigerator	1003	$121
Powered down empty 37°C Shell lab shaking incubatore	~900	~$100
Powered down empty 35°C incubator	752	$91
4 heat blocks timed off during nights and weekends	722	$87
Elected not to replace newly broken −20°C GE freezer	672	$81
Powered down Thelco 130D Precision Scientific Lab ovenf	~650	~$80
Surplussed LRP fridge/freezer combo	499	$60
Surplussed old Mac computer	263	$32
Powered down 42°C gel-prep water bath	84	$10
Estimated total	~83000 kWh	~$10,000

^aAscribed savings from interpolated data (12) that were averaged with measurements of kWh consumed by one ULT freezer in study at two set points, for more information see methods.

^bUnable to direct measure autoclave – consumption estimated from My Green Lab^® report (13).

^cUnable to direct measure BSCs – consumption estimated from CU Boulder estimates (22).

^dSavings are the difference in energy consumption between the old and new refrigerators.

^eOutlet incompatible with energy monitor, applied average measurement from two similar sized and aged incubators.

^fPlug incompatible with energy monitor, value estimated from similarly sized Thermo Forma 310 Direct Heat CO₂ Incubator.

Our findings indicate that increasing ULT units to −70°C (N=30) and −75°C (N=5) likely resulted in the largest estimated gross annual savings (~40,500 kWh; $4,900), followed by powering down a steam-jacketed autoclave (13) 79% of the week (~23,400 kWh; $2,800), and powering down two BSCs after 5 p.m. and over weekends (~8,300 kWh; $1,000). However, aside from measurement of one ULT at two setpoints, values for ULT freezers, BSCs and the autoclave were estimated based on previous reports. Using previous data, the largest impact per device in our case study occurred by powering down the autoclave, then BSCs. Other large savings were obtained by replacing old, inefficient, sliding-door fridges with previously purchased efficient models and powering down an old, large shaking chest incubator.

Powering down the older BSCs and autoclave in this study relied on user-compliance. One benefit of users taking action to power down devices is that it encourages knock-on effects such as habitual alteration in other daily tasks (e.g., consideration of the environment); however, due to incomplete adherence, this may also result in devices accidentally staying on when users forget. Updated models of devices often offer energy-saving measures, especially for large consumers such as autoclaves. For example, the brand of autoclave in this study (Consolidated sterilizer systems, Billerica, MA, USA) currently offers new sustainability features unavailable in our older models including calendar-based start-up/shut-down features, an “EcoJacket” to decrease the jacket temperature between cycles, automatic shut-off during idle time, and water savings. Options to retrofit older models in circulation would provide a rich opportunity to save energy and water. Based on the My Green Lab^® autoclave report (13), powering down a single front-loading steam-jacketed autoclave after hours saves roughly 26 billion gallons of water annually. Along these same lines, the company responsible for maintaining the BSCs in this study recommended an older model never be turned off, as this led to issues powering-on. Unfortunately, purchase of a new BSC was prohibitively expensive. Future work to coordinate between suppliers and University administrators to offset initial costs with long-term benefits would benefit both parties.

This case study involved labs operating on an open floor plan with ~100 employees and 19 principal investigators. Extensive education, coordination, and collective agreement was therefore required to ensure successes. The advantage of this model is the effects of interventions are magnified. Furthermore, peer pressure involved in shared space can encourage compliance.

Cold-box cleanout

The cold-box cleanout day resulted in 1) donation of five refrigerators (two fridge/freezer combo), and 2) the eventual choice not to replace a broken −20°C unit. This equated to direct measurement savings of 6,489 kWh and ~$782 saved per annum, which represents 8% of total savings. Importantly, safety and accessibility were improved by reduction of hallway blind spots, and physical space was freed for incoming investigators.

Due to the large number of longstanding investigators on our floor, equipment is numerous and often redundant (e.g., roughly 40 incubators). Parting with, and sharing, equipment can be a difficult decision. However, shared equipment programs can benefit institutions for extensive reasons (25), not least of which is the responsible use of funding to research sponsors. A shared instrumentation program at CU Boulder saved ~$3 million in avoided equipment purchases from 2018–2022 with the program estimating $4.5 million/year in savings if the program were expanded campuswide (26). Identifying roadblocks to scaling up such implementations is critical.

ULT freezer campaign projections

Our freezer campaign (Fig. 2a) resulted in increased temperature setpoints for 35 ULT units owned by 10 PIs. We estimate annual savings to be ~40,500 kWh, equivalent to over four midwestern homes, Table 1. If adopted university-wide, projected annual savings are estimated to be 875,000 kWh, or ~92 midwestern homes, Fig. 2b. Annual financial savings for this simple, one-time intervention were estimated to be ~$4,900, while projected University-wide savings are estimated at >$105,000, Fig. 2c, equivalent to over two years of indirect costs (at 35%) of an R01 equivalent-level award based on 2023 funding levels (27), assuming a 4.5-year award length. Savings from temperature increases are inversely related to ULT age. Because 49% of the University’s 702 ULT units were purchased in 2015 or prior (and 190 from 2010 or prior), savings are likely an underestimate.

Fig. 2. Estimated energy savings from ULT freezer campaign.

Savings are an average of values reported in (12), and direct measurement of one ULT freezer. a. Signage posted in common areas and at the entrance to an ULT freezer farm; b. Annual estimated energy savings (kWh) from ULT freezer campaign amongst collaborating PIs (N=10) and projected energy savings if interventions were scaled University-wide (based on 2023 university inventory report); text above bars represents kWh savings converted to the number of equivalent Midwest homes; c. Estimated annual financial savings of ULT temperature increases, and financial savings estimates if scaled University-wide; text above bar represents equivalent years of indirect costs (at 35%) from one R01-equivalent grant.

Values from the ULT freezer intervention are best estimates. We provide a gross estimate of energy savings from ULT units using a combination of available data in the Center for Energy Efficient Laboratories report from California (12), representing slightly older units (N=7), and our own direct measurement of a newer unit at both set-points. The average volume of representative units was 503L. We had difficulty accessing all unit plugs, incompatibility with the watt monitor on some devices, and found significant heterogeneity in age of the 35 units. Additionally, coordination with 10 PIs to allow us to measure each device was beyond the scope of this work. However, the goal of this calculation was to give an estimate of overall energy savings from a wide variety of ULT freezers, and not to compare differences between freezers.

Savings are likely underestimated. As floor-wide initiatives became normalized and power-down stickers ubiquitous, individual actions were not always captured. For devices with higher voltage requirements, savings were interpolated based on best-estimates from similarly functioning and sized devices. However, these devices were inherently more efficient given their lower voltage requirements, and newer age. ULT temperature energy consumption was only directly measured for one unit and data interpolated from (12) to estimate consumption at both set-points. An additional resource for researchers seeking information about ULT freezer consumption is the Supplemental Table 36 of (17). Similarly, BSC consumption was estimated from previously available resources (9,19). This likely led to underestimation of energy savings from specific interventions. University scale-up estimates (Fig. 2b) are based on the University-reported ULT units and do not include those used by university professors working at the Veteran’s Affairs center, which are numerous.

Many universities in the U.S. have impressive Green Labs programs (26,28) and are making tremendous advances towards resource efficiency. This work is often reported in online resources (13,22,25) and through professional networks. Comprehensive work describing the gaps in sustainable action in scientific endeavors (29) and lab spaces (16,18,30,31) has been reported in scientific journals, typically from groups in Europe and the United Kingdom. This often voluntary work (17) can also lead to significant savings. For example, Freese et al. (2024) demonstrated the University of Groningen saved nearly 500 tonnes of CO₂e and 400,000 euros by voluntary efficiency measures amongst 46 laboratories (17). Some of the same investigators concomitantly published a living guidebook for investigators to improve sustainability in their lab (30). This extensive document covers topics ranging from greener solvents to dry-lab etiquette and provides useful stickers to nudge others to action. Although some lab equipment specifications can vary slightly in the U.S. (e.g. fume hood model), the principles of this work are universally relevant and provide a helpful resource for labs interested in beginning a sustainability journey.

Previous research sustainability work has found grassroots efforts to improve sustainability in research are hampered by time and budgetary constraints, and management-related involvement (16). Durgan et al. pointed out that while researchers are accustomed to acting after expertise has been achieved, one need not be an expert to begin improving sustainability in the lab. They also point out that it helps to build a team, organize and involve senior staff (18). The tools scientists develop in bench work and data analyses are useful and appropriate toward cataloging sustainability progress. We found certain steps are easier to implement than others, for example, turning up ultralow temperature freezers. Investigators were amenable to single requests, which are easy to sustain compared with group behavioral changes. Although these actions are additive, as previously noted (32), the National Institutes of Health does not currently require labs to become externally certified (33,34) for sustainability features, a roadblock to widespread adoption in the United States.

Our efforts resulted in divisional buy-in to fund My Green Lab^® certification, and were promoted within the department (35). Despite this, institutional support and expansion remains a largely unmet need. For example, real-time energy monitoring of each campus building was historically provided as a free service but is no longer available. By 2024 this “energy dashboard,” was not functional and the University of Iowa utilities had been purchased by French multinational electric utility company, ENGIE by public-private-partnership. While ENGIE is working with the BSI to provide energy estimates, real-time open-access tools like the energy dashboard are deeply important for movements such as ours to gather data, improve indirect funding efficiency, and monitor successes. Accordingly, we provided access to our database to fellow researchers. Fortunately, the University has developed an Asset Optimization Services team performing vital work. The team is approachable, and has access to building level data connecting campus buildings to a fault detection monitoring system, saving the university $600,000 in the first six months (36).

Conclusions and Recommendations

Now more than ever, universities are incentivized to save money. Because research laboratories are some of largest consumers of resources on campus, they are one of the best targets for administrators looking to save indirect funding. In our case study at a U.S. public university, actions to fine-tune temperatures and optimize use of devices operating at large deviations from room-temperature we estimate led to the largest energy savings, while per-device, powering down an autoclave resulted in the largest savings. Increasing the temperature of ULT units is simple and effective, with little required maintenance. As has previously been reported, other important targets are devices designed to constantly pull filtered air through air-handling systems, such as powering down BSCs and “shutting the sash” on fume hoods. Full-time staff dedicated to improving efficiencies in research should garner universities the largest gains, however, individual investigators can still save universities energy and money. We estimate that our small, unfunded initiative saved the equivalent energy of over nine average Midwest homes, while improving efficiency of indirect funds. Based on previous work as well as our experience, we recommend that others hoping to get involved in green labs begin by presenting background facts to their labs. Then finding other interested staff, organizing/team building and collecting data on current and projected energy consumption. Based on collected data we then recommend taking, or recommending, action to reduce consumption of the largest consumers. Ideally this process will involve senior staff and faculty. On an administrative level, universities will save the most resources if they fund full-time “green lab” staff and promote external sustainability certification.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its supplementary information files. Original Microsoft Excel files are available from the corresponding author on reasonable request.