presents
Leaking Havoc
Exposing Your Supermarket’s Invisible Climate Pollution
Exposing Your Supermarket’s Invisible Climate Pollution
presents
Leaking Havoc
Exposing Your Supermarket’s Invisible Climate Pollution
The investigation focused on Walmart and other top-grossing supermarkets in the region, documenting leaking systems and capturing invisible climate pollutants on video. Sixty percent of the Walmart stores investigated by EIA had a refrigerant leak. Fifty-five percent of the stores we visited were measurably leaking super pollutants. The invisible nature of the gases has allowed companies to overlook refrigerant emissions and calls to action; but with the right equipment, these gases are neither invisible nor immeasurable.
Refrigerant leaks were detected in 60% of the Walmart stores investigated, meaning over half of all stores were leaking.
While Walmart corporate sustainability reports claim that responsibly managing its refrigerant use is a high priority, our on-the-ground detection raises serious concerns about potential widespread leakage at Walmart stores.
Leaks persisted months later in the few stores revisited – an indication that routine leak inspection or repair did not occur, or was not effective in detecting the leaks.
Across all other companies visited, 55% of the stores we visited were measurably leaking, consistent with concerns that refrigerant leaks are an industry-wide problem.
Both company policies and government regulations need improvement to ensure all leaks, even those detectable only at small concentrations, are promptly identified and repaired in every store.
EIA findings scratch the surface of the systemic leakage issue, as a majority of refrigerant leaks occur away from the display cases in areas inaccessible to everyday customers and our investigators. The problem is likely much worse.
Hydrofluorocarbon (HFC) refrigerants account for 65% of a food retailer’s scope 1 greenhouse gas emissions on average.1
Proactive refrigerant management, such as routine inspection and leak detection equipment installation to avoid leaks, and the elimination of super pollutant use could significantly reduce the climate impact of supermarket operations. In 2020, major supermarket chains experienced skyrocketing profits; on average, major grocers netted an extra $16.7 billion USD in earnings compared to 2019.2 Still, many of the largest supermarkets are yet to institute meaningful and systemic changes to address super pollutant leaks. Walmart, the largest retailer, has failed to implement voluntary commitments to reduce refrigeration emissions that the company made more than a decade ago and delayed transition to climate-friendly alternatives.
What's a Refrigerant?
All cooling systems, from a small window air conditioner to a large supermarket refrigerator, use some type of coolant, or refrigerant. The most common refrigerants you will find in the market today are HFCs. HFCs were introduced in the 1990s as a replacement for ozone depleting substances (ODS), which are being eliminated globally under the Montreal Protocol.3 While HFCs have a negligible impact on the ozone layer, they are many thousands of times worse for the climate than CO2. Pound for pound, HFCs trap substantially more heat than the most common greenhouse gases, with global warming potentials (GWPs) as high as 12,690 – rightfully earning their title of “super pollutants.” 4
HFCs are projected to comprise 19% of total global greenhouse emissions by 2050; addressing these gases has been recognized as the single largest climate mitigation opportunity available.5
Under the Kigali Amendment to the Montreal Protocol, the global community is already working to phase-down HFCs and transition to climate-friendly alternatives. Alternative refrigerants with ultra-low GWPs such as CO2, propane, ammonia, and other hydrocarbon blends are already being used in commercial applications today.
How Much Supermarkets Leak
Supermarkets, with their massive refrigeration demands, are among the largest corporate users of HFCs and the remaining bank of ODS refrigerants. A single store contains an average of 3,500 pounds of HFCs, and many more thousands throughout their entire cold chains, which includes transportation refrigeration units, refrigerated shipping containers, cold storage warehouses, and in-store refrigeration. In the United States, there are approximately 38,000 supermarkets and only a few hundred are completely HFC-free in their central refrigeration.6
Massive commercial refrigeration systems are highly susceptible to loose components and refrigerant leakages. These systems have lots of complex piping. Any improper installation, maintenance or tightening can result in loose or broken fittings which weaken connection points and allow for leaks.7 The mechanics of a refrigerator, which operates by expanding and contracting refrigerant under pressure, adversely affects the integrity of seals.8 In fact, vibration and regular wear and tear itself can exacerbate leakages. For these reasons, it is nearly impossible to have a system that will never leak in its lifetime.
On average, a U.S. supermarket leaks 25% of its refrigerant charge a year, harming our environment as well as corporate checkbooks.9 When these high-GWP gases, and in some instances still ODS, inevitably leak, they pose a direct threat to our climate and environment. One store using HFCs leaks the equivalent of more than 300 additional cars on the road a year.10 Across the sector, the emissions from leaking refrigerants in the U.S. alone is equal to 49 billion pounds of coal burned in a year.
Click to reveal the climate impact of supermarket refrigeration across one store, one company, and one country:
Managing the refrigerant that is used is essential to reduce the climate impact of these systems; an even more effective solution, however, is to discontinue the use of high-GWP gases altogether. Commercial refrigeration is one of the most advanced cooling sectors when it comes to adopting climate-friendly refrigerant solutions. The barriers to adopting better, climate-friendly technology in this sector are less pronounced than in other cooling applications. In fact, in Europe ultra-low GWP natural refrigerants are the new standard for supermarket builds, with over 26,000 supermarkets using transcritical CO2 technology, not to mention those using hydrocarbons, ammonia, or other system types.11 This is in stark contrast to the barely 600 known stores in the U.S. using the same technology.12
We found leaks using advanced leak detection technologies.
"If you encounter any significant gas – 2 ppm or more – I would strongly suggest that you start looking in earnest for what might be a major leak in the case. Always keep in mind that leaks are going to be at dispersion. So, 2 parts per million in the front of this case could be a much greater leak as the air mixes and disperses throughout the unit. So any gas at all is an indicator that you have some kind of leak in process"
– Instructional Video by Bacharach, Manufacturer of PGM-IR Portable Leak Detector13
EIA using leak detectors in supermarkets
To detect leaks we used three different advanced leak detection tools:
Bacharach PGM-IR Bagless Portable Refrigerant Monitor: capable of detecting over 60 different refrigerants including various chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), and HFCs.14 The device is refrigerant specific and requires user selection of the refrigerant of interest. The machine compares the infrared (IR) signatures of the intake air to the preprogrammed signature for the selected refrigerant. The device has an accuracy of +/-1 ppm for most HFCs and +/-10 or +/-15% for HCFC-22.15
Inficon D-TEK Stratus Refrigerant Leak Detector and Portable Monitor: capable of detecting all CFCs, HCFCs, HFCs, hydrofluoroolefins, blends and CO2.16 The Inficon D-TEK Stratus does not require the user to know the refrigerant of interest. Instead, the device references a library of IR signatures and outputs a ppm value if the intake matches any of the signatures in the reference library. The device has an accuracy of +/-1 ppm or 10% of its reading.17
FLIR GF304 Refrigerant Leak Detection Camera: assesses the absorption and reflection of thermal wavelengths to visually detect HFCs.18 The camera filters the infrared spectrum to detect spectra in a small range: 8.0-8.6 micrometers. This is the spectral range that HFCs absorb. The camera’s high-sensitivity mode allows for leak verification, as not only is it filtering for the specific spectrum but also it is simultaneously accounting for any motion, thereby emphasizing the contours of movement, which is useful when detecting fleeting gases.
The Bacharach PGM-IR Bagless Portable Refrigerant Monitor was the primary device used. The Inficon D-TEK Stratus was used to corroborate the findings of the Bacharach when possible, and the FLIR infrared camera was used as a complimentary tool to visualize leaks and their dispersion. Using different technologies helped us further validate our findings, and address any potential gaps in data.
Following each device’s unique handling and use instructions, we entered every store with the Bacharach PGM-IR machine turned on and warmed up outside of the store. When inside, the device was programmed to detect the refrigerant labelled on the refrigerated cases inside the store. The procedure laid out in Bacharach’s “PGM-IR Detecting Grocery Store Leaks – Sales floor” video was followed, consisting of checking low to the ground in cases and at every other door and grate opportunity.19 To determine whether there was a leak from the central refrigeration system in a given store, we followed this technique and recorded a positive leak if the machine read greater than 2 ppm. When detecting R22 we only considered a positive leak if it was greater than 10 ppm as the margin of error for detection of this refrigerant is higher than most other refrigerants. In most cases, the presence of a leak was also confirmed by the replicability and cross reference between both the Bacharach and Inficon devices.
The ppm recorded for each store (see figure 2) is the highest ppm the leak detection equipment observed at that point in space and time. This number does not quantify the size and severity of a leak, which may vary based on the timing and location of the leak detected and the amount of dispersion. Since we are not measuring the severity of a leak, we took a binary approach to recording the presence of a leak or no leak. For this reason we are not sharing the ppms recorded in each store. However, all leaks recorded using this procedure had the potential to be major leaks, even if detected at a small concentration of <10 ppm, given the likelihood of considerable dispersion throughout the large space.
Since we were limited in our ability to further investigate the leak once detected, we made other considerations while detecting to ensure the accuracy of our results, including:
While this is a limitation of our methodology, it also suggests that our study undercounted supermarket HFC leaks. A more thorough investigation, with access to more than just the sales floor, may show that supermarket leaks are even more common and severe than we report.
As soon as we were inside one Giant store, the Bacharach PGM-IR began to report a background reading of 4-5 ppm of R404a, which has a GWP of 3,922. This occurred both in closed cases and just strolling through unrefrigerated aisles. Clearly there was a leak, but where? Further investigation suggested the leak was coming from a closed freezer case, where we recorded readings of about 12 ppm. We continued down an open refrigerator aisle without doors on the display cases. Suddenly, the Bacharach PGM-IR started reporting a much higher concentration of 80-90 ppm, with a peak of 92 ppm. We were then able to use the FLIR camera to visually capture the leak spewing from the top of the refrigerated case and out over the products it contained. The source of the leak, at the top of an open case, made this leak easy to capture on camera, more so than a leak originating inside or at the bottom of a refrigerated case. To our knowledge, this is the first time that a display floor refrigerant leak has been caught on camera, bringing this invisible climate killer into plain view.
The FLIR GF304 is only sensitive to infrared light in the small wavelength band which HFCs absorb. For that reason, no infrared light is reflected by HFCs back into the camera lens and so they appear black on camera. What resembles black smoke in this footage is actually HFC gas. The change in video quality to higher contrast footage is when the camera was set to “high-sensitivity mode” which accentuates movement as well as refrigerant gases.
Walmart is the largest retailer as well as the largest food retailer in the U.S.22 In 2019, Walmart grossed $399.8 billion USD in U.S. retail sales which is over double that of the next largest retailer.23 In the U.S., the company operates 4,748 Walmart store locations and 599 Sam’s Clubs.24 Globally, Walmart operates 11,510 locations under 55 banners, with a U.S. store growth rate of 1.7% between 2018 and 2019, and a U.S. sales growth rate of 2.6% over that same period.25 The company’s size and global presence uniquely positions it to transform the U.S. refrigeration market. Walmart self-reported HFC emissions of 2,661,801 MTCO2e in 2019, that is the same climate impact as driving from Seattle to Miami two million times.26 Nevertheless, there is little publicly available information about Walmart’s actions to minimize its use and emissions of HFCs. Meanwhile its corporate sustainability reports claim that responsibly managing refrigerant use is a high priority.
Of the 20 Walmart locations visited, 60% had a confirmed leak – meaning over half of all Walmart locations visited are leaking greenhouse gases that wreak havoc on our climate.
Concentrations of gas found on the sales floor ranged from as little as 2 ppm to 182 ppm. In our sample, Walmart used three types of labelled high-GWP refrigerants, R407A, R404A and R22, with global warming potentials of 2,107; 3,922; and 1,810, respectively.
Over the years, Walmart has made a number of statements on plans to address the impact of its refrigerants. In 2010, Walmart along with other leading retailers, committed under the Consumer Goods Forum Refrigerant Resolution to begin phasing out HFCs starting in 2015.27 This commitment included transitioning to HFC-free refrigeration in new and refurbished stores.
A decade later, Walmart does not have a single store in the U.S. using HFC-free refrigeration in its central systems, nor has the company instituted a comprehensive HFC policy and management strategy.
Its latest Environmental, Social, and Governance Report does not include any refrigerant related goals or details on any HFC reduction efforts.28
In September 2020, Walmart announced it would be transitioning to “low-impact refrigerants” for cooling and electrified equipment for heating in its stores, clubs, and data and distribution centers by 2040, a deadline that puts them decades behind some major competitors.29 In this announcement, “low-impact refrigerants,” are not defined, nor are any intermediary goals over the next two decades to achieve these objectives. Walmart also failed to offer a plan to address current leaks of HFCs from its stores in its announcement.
Case Studies
Select a case study to read more.
More than half of all stores investigated had refrigerant leaks detectable on the sales floor.
Figure 1: Percent of Walmart and all other stores with a measurable leak.
It remains legal to leak large quantities of refrigerants. Federal leak regulations currently omit HFCs and have lenient thresholds for action.
Significant gaps remain in U.S. regulations and enforcement to reduce refrigerant emissions, however recent legislation, yet to be implemented, provides an opportunity to correct course. The second COVID-19 stimulus package, passed in December 2020, included bipartisan HFC legislation called the American Innovation and Manufacturing (AIM) Act, authorizing the Environmental Protection Agency (EPA) to implement additional regulations to phase down HFC production and use consistent with the Kigali Amendment as well as regulations to minimize releases during the lifecycle of equipment use, maintenance, and disposal.32 The inclusion of an HFC phase-down in this bill represents the most significant climate legislation passed by Congress in over a decade and sends a clear message to the market that HFCs need to be addressed this decade.
The omission of HFCs from leak management requirements is unpopular amongst the HVAC industry.
Section 6 of the Clean Air Act (CAA) focuses on ODS and their substitutes, including HFCs. Section 608 in particular establishes the National Recycling and Emission Reduction Program that among other things, regulates and minimizes leaking refrigerants. The Program does this by requiring owners and operators of large refrigeration and air conditioning systems, including supermarkets, to undertake certain activities to track and repair chronic leaks. EPA performs random inspections, responds to tips, and pursues potential cases against violators of the Section 608 regulations. EPA is also authorized to assess fines for violations of these regulations.
Under the Obama administration, EPA extended Section 608 of the CAA refrigerant management requirements to cover HFCs as well as ODS.33 Under this Section 608 provision, supermarkets containing over 50lbs of refrigerant would be required to quantify and keep record of leaks. Only if equipment leaks over a threshold of 25% of its charge annually are additional requirements triggered to repair leaks, conduct regular inspections, and for some of the worst leaking systems with repeated leaks, retrofit or retire and replace the system entirely.
In 2020, the Trump administration rolled back the Section 608 provisions to omit HFCs from leak management requirements, citing insufficient authority, leaving current leak repair requirements only covering ODS.34 However, California and several other states have moved to expand state-level regulations on HFC leaks, while the AIM Act provides new authority to EPA. The rollback of Section 608 provisions must be immediately restored and improved upon under EPA’s new authority. More robust controls on leaks should require regular preventative detection in all stores, capable of alarming and triggering inspection and repair of all detectable amounts of HFCs or ODS, even at low ppm concentrations. All repairs should be conducted within a strict timeframe, with robust recordkeeping and associated enforcement and penalties for violations.
Examples of Federal Enforcement on Non-compliance
To this day, a significant portion of leaked refrigerants are ODS. Still undergoing phase-out under the Montreal Protocol, ODS continue to be legally used in equipment installed prior to the phase-out date. ODS are thus still used in commercial refrigeration systems, that are prone to leaking. In the past decade, four unique federal enforcement actions have charged supermarkets for violating Section 608 CAA requirements and leaking ODS. In all four instances, major food retailers either failed to promptly detect and repair leaks above the leak threshold and/or violated recordkeeping requirements. Enforcement actions against offenders include:
Supermarkets must urgently act to stop these unnecessary climate pollutants from leaking into our atmosphere.
While individual corporate action is important, government policy is necessary for industry-wide change. EIA calls on federal and state regulators to enact policy solutions to reduce leaks and end-of-life emissions from supermarkets, including:
This report primarily serves as a call to action to industry and government to act in the interest of our shared climate. It also demonstrates the feasibility of finding leaks in your neighborhood and serves as a tool to hold the largest companies accountable. You can:
Our results only scratch the surface of the widespread leakages across the industry.