CHAPTER FOUR
FINDINGS
Google’s Energy Terminology
My examination of Google’s public-facing documents and materials revealed explicit discourse related to energy buying and matching, power purchase agreements (PPAs), green tariffs, renewable energy credits (RECs), power usage effectiveness (PUE), and carbon neutrality. An important step in critical discourse analysis is to also contextualize such discourse, which involves analysis of documents from others, including government agencies. In 2007, the Environmental Protection Agency (EPA) released a report about energy use in the country’s data centers, a “vision for achieving energy efficiency in U.S. data centers” (p. 13). The EPA’s report also states that, “Although the growing energy use of servers and data centers makes this a challenging goal, there are large opportunities for savings. These savings will not be easy to achieve, given the barriers outlined in this report, but there are many policies available to overcome the barriers” (EPA, 2007, p. 13). The documents analyzed contain Google’s explanations about such barriers they face as a company.
It is important to note that Google is not the only one promoting their core values, goals, and commitments to clean energy. In 2019, for example, The Environmental Protection Agency (EPA) selected Google as their “Green Power Partner of the Year” Award, which is a category that “is the highest organizational honor in EPA’s Green Power Leadership Awards and the activities are commensurate with this level of recognition” and “recognizes Partners that distinguish themselves through their green power use, leadership, overall strategy, and impact on the green power market” (EPA, 2019). The EPA continues, stating that the winners of this award “represent a beacon for other organizations to follow, represent best in class in terms of market impact, and have a compelling story that is both unique and replicable to a wider set of market participants” (EPA, 2019). Google is also a founding member of the Renewable Energy Buyers Association (REBA), whose “goal is to catalyze 60 gigawatts (GW) of new renewable energy projects by 2025 and to unlock the energy market for all large-scale energy buyers by creating viable pathways to procurement” (REBA, 2020). Google leadership, including Mitchell Terrell, Google’s Head of Market Energy Development, are members on the REBA board, alongside company executives from Amazon, Salesforce, Facebook, General Motors, and more. Such context bolsters the following claims and positions Google promotes, also situating the company as one of the leaders in the renewable energy marketplace. Such alliances, like REBA, matter to Google because, as Ruth Porat, the Senior Vice President and Chief Financial Officer at Google, wrote: “We firmly believe that every business has the opportunity and obligation to protect our planet” (Porat, 2019).
Sustainability as Value
Google has positioned itself as both invested in and conscious of the relationships they have with the resources required to make their operations function: “At Google, we care about energy for many reasons, but fundamentally it’s because our business depends on it” (Google, 2016a, p. 2); and “creating a carbon-free future will be no easy feat, but the urgency of climate change demands bold solutions”(Terrell, 2018, para. 6). Google’s documents express how the company views these relationships as “commitments.” These commitments matter to Google because, “electricity is the fuel that allows our data centers to deliver billions of Google searches, YouTube views, and much more—every single day, around the clock. Our commitment to carbon-free energy should be around the clock too” (Terrell, 2018, para. 1). Beyond their business, “Google is committed to being part of the solution to solving global climate change, both through purchasing renewable energy to match the energy use of our own operations and by helping to create pathways for others to purchase clean energy themselves” (Google, 2016a, p. 3). This disclosed commitment to being part of the climate change solution goes back to Google’s foundational, core values:
“Operating our business in an environmentally sustainable way has been a core value from the beginning, and we’re always working on new ideas to make sustainability a reality — like enabling the building of healthy workplaces and creating a living, breathing dashboard for the planet. We’ve reported our carbon footprint and published information on our sustainability programs for many years in white papers, blog posts, and on our website” (Hölzle, 2016, para. 7).
Again, other documents included explicit discourse, reiterating such business commitments: “At Google, our values reflect the fundamental importance of inclusion, openness, science, and commitment to the environment. Operating our business in an environmentally sustainable way has been a core value from the beginning,” (Google, 2016b, p. 1); “Google has a longstanding commitment to climate action and environmental stewardship. Sustainability has been a core value since Google’s founding, and we strive to build sustainability into everything we do” (Google, 2019, p. 2).
Other documents indicated when Google made strides towards accomplishing their goals: “In 2017, Google achieved a major milestone: purchasing 100% renewable energy to match our annual electricity consumption for global operations, including our data centers and offices. We did it again in 2018” (Google Data Centers, n.d., para. 1). Such explicit mentions related to commitments continued across the documents in this project and appear as another exemplification of Google’s purported pioneering on the sustainability front. Through each of the quotations pulled, Google articulated their relationships to sustainability and carbon neutrality as a “commitment” or “value.”
Sustainability as Commitment
In 2016, Google released an Environmental Report that comprehensively outlined the steps the company has taken and plans to take to be more sustainable in operations. Urs Hölzle, Google’s Senior Vice President of Technical Infrastructure, opened the document expressing how “our values reflect the fundamental importance of inclusion, openness, science, and commitment to the environment” (Google, 2016b, p. 01). The word “commitment” appeared 18 times in this document alone, with each use referring to a form of financial or contractual agreement that builds off the promise to better the environment through Google’s practices (Google, 2016b). Nearly 3 years later, the 2019 Environmental Report began with Urs Hölzle and Ruth Porat, Google’s Senior Vice President and Chief Financial Officer, expressing that “Google has a longstanding commitment to climate action and environmental stewardship” and “Sustainability has been a core value since Google’s founding, and we strive to build sustainability into everything we do” (Google, 2019, p. 2). In the 2019 report, the word “commitment” appeared 22 times and implies promises, financial agreements, and company stances (Google, 2019). As my analysis continued, commitment remained a key term with such implied meanings, often reiterated by company leadership, and I believe it is worth noting here because it is central to how they frame their corporate role in the climate crisis. A commitment can be a value, a financial business agreement, and a stance.
Google and Renewable Energy
To further understand what it means for a company like Google to have an obligation or commitment to “build sustainability into everything” they do (Google, 2019, p. 6), Google posed a question: “Let’s say it’s 2009. You’re a global technology company, and you need very large amounts of renewable energy. How do you get it?” (Google Sustainability, n.d., para. 1). Documents announced how Google’s “large-scale procurement of wind and solar power is a cornerstone of our sustainability efforts, and has made Google the world’s largest corporate buyer of renewable energy” (Google Data Centers, n.d.-b, para.1). In the last few years, Google, somewhat redundantly, touts big accomplishments in execution of their commitments: “In 2017, we became the first company of our size to match our entire annual electricity consumption with renewable energy (and then we did it again in 2018). As a result, we became the largest corporate buyer of renewable energy in the world.” (Pichai, 2019, para. 2). Documents indicated that such accomplishments occurred through their purchasing prowess:
In 2012, Google made a commitment to purchase enough renewable energy to match 100% of our operations, and we are excited to announce that we will reach that goal in 2017. Reaching our 100% renewable purchasing goal means that Google will buy on an annual basis the same amount of megawatt-hours (MWh) of renewable energy—both the physical energy and its corresponding renewable energy certificates (REC)—as the amount of MWh of electricity that we consume for our operations around the world. (Google, 2016a, p. 1)
In a 2017 report from Alphabet, Google’s parent company, climate change is described as integrated into business strategy through renewable energy. Namely, Alphabet disclosed that in order to “mitigate future price rises in electricity costs” they “seek long-term contracts for renewable electricity” and are “relentlessly focused on improving energy efficiency in our facilities, including data centers and office spaces” (Alphabet, 2017, p. 8). Renewable energy is important to Google’s corporate climate change strategy.
Not only do the documents repeat one another reiterating the company’s commitments and the purchases they have made to reach them, they sometimes break down the process. For example, in 2010, a post on Google’s official blog by Urs Hölzle, Senior Vice President of Technical Infrastructure, detailed the company’s step-by-step processes: “First, we minimize our energy consumption; in fact, we’ve built some of the world’s most energy efficient data centers. Second, we seek to power our facilities with renewable energy, like we did in Mountain View, CA with one of the largest corporate solar installations. Finally, we purchase carbon offsets for the emissions we cannot directly eliminate” (Hölzle, 2010, para. 1). In 2016, Google representatives gave more details related to their processes: “We have aggressively employed a variety of purchasing tactics, some of which are closer than others to our ultimate goal to supply our operations with 24-7 clean energy. We use four primary tactics: 1. “Direct” renewable purchasing, 2. “Offsetting” renewable PPAs (aka “fixed-floating swaps”), 3. Utility renewable energy tariffs, 4. Grid-mix renewable content” (Google, 2016a, p. 5). Throughout these documents, multiple, perhaps unfamiliar, terms have been presented and will be explored to understand how the company explains their accomplishments. The following sections analyze the discourse related to explaining the purpose and process of renewable energy acquisitions.
Carbon Neutrality
Google needs energy, wants to buy renewable energy, and plans to minimize their carbon footprint on a level and rate they claim to be unmatched by their fellow competitors and colleagues. Carbon neutrality, or essentially claims that the carbon added into the atmosphere is matched by carbon detracted from the atmosphere, is something Google boasts about: “Google’s cloud services are all carbon neutral—dating back to 2007” (Google, 2011, p.1). Google understands carbon neutrality and greener operations are not isolated to energy though that is a very important, prominent facet of their work. Instead, documents also reveal how Google hopes to “maximize the reuse of finite resources across our operations, products and supply chains and enable others to do the same” (Brandt, 2019, para. 8). Google uses the term “carbon neutrality” alongside mentions of renewable energy; “We depend upon large quantities of electricity to power Google services and want to make large actions to support renewable energy. As we continue operating with the most energy efficient data centers and working to be carbon neutral, we’re happy to also be directly purchasing energy from renewable resources” (Hölzle, 2010, para. 4). Google’s support for renewable energy comes through adding renewable energy projects to their portfolio with big implications; “Once all these projects come online, our carbon-free energy portfolio will produce more electricity than places like Washington D.C. or entire countries like Lithuania or Uruguay use each year” (Pichai, 2019, para. 3).
Buying Energy
To acquire renewable energy, Google representatives wrote that, “Ideally you would just buy it from your local utility. But you can’t, at least not yet: Most utilities are still heavily regulated entities whose business model — keep the lights on and prices reasonable — lacks both mechanisms and incentives to respond to customer requests for renewable energy” (Google Sustainability, n.d., para. 2). To overcome this and when discussing such renewable acquisitions, a popular term Google uses throughout their materials is Power Purchase Agreements (PPAs). They describe a PPA as “a contract to buy power over a period of time at a negotiated price from a particular facility” (Google, 2013, p. 3). The company became interested in PPAs over a decade ago and described the process: “In 2009, our data center energy team began to study power purchase agreements (PPAs): large-scale, long-term contracts to buy renewable energy in volumes that would meet the needs of our business (Google Sustainability, n.d., para. 5).” Google explains that they cannot buy enough “clean energy” from the utility companies on the grids of their infrastructure, and due to geographical limitations, they are unable to produce enough of their own green energy on site (Google Sustainability, n.d.). Instead, they say that, through wholesale agreements, they can purchase green energy from developers who operate on the same power grids as their data centers (Google Sustainability, n.d.). Essentially, Google has contracts with renewable projects that are specifically located on the same power grids as their data centers so that, since they cannot “legally or physically” transfer that power directly to their centers, they can access that energy through the grid (Google, 2013, p. 3). It is important to understand that energy, or the individual comprising electrons, exist in ways that make it impossible to differentiate between which were created via green or “brown” sources. Hypothetically, PPAs could also be thought of as adding drops of water into a pool; the particles in the pool will be changed with each addition, making it possible to extract water droplets with varying concentrations of whatever was added. One may not know where such additions began or end, but the entire body of water is changing in composition. Because of this, when Google adds their green electrons to the grids through PPAs, the result is a process that also helps to green the regional power grids overall. Google representatives stated that, “while the renewable facility output is not being used directly to power a Google data center, the PPA arrangement assures that additional renewable generation sufficient to power the data center came on line in the area” (Demasi, 2013, p. 2). This is how PPAs work- Google purchases green electrons and adds them to the local power grids, making them available for use and in essence, changing the composition of the entire pool of energy generated to something purportedly “greener.” Legally, such dilutions count towards the greenness of the energy used by the organization who owns the purchases.
Though Google does hope to buy renewable energy from projects located within the same power grids as their data centers, sometimes this is not always possible. There is a difference between buying energy and “matching” the energy used all hours of the day. In fact, one document acknowledged this difference, explaining how Google matches their “annual electricity consumption,” which does “not mean that our facilities are matched with renewable energy every hour” of the day (Google Data Centers, n.d.-b, para. 2). The same document also expresses the impossibility and the ways they overcome sourcing enough green energy on their properties: “to compensate for times and places in which the wind slows or sunlight fades, we currently buy a surplus of renewable energy at other times and in other places” to accomplish their 24/7 carbon neutrality (Google Data Centers, n.d.-b, para. 2). Google’s discourse explained how, ultimately, they are interested in ensuring their renewable energy buying is at least equal to their energy consumption, ending at a net zero or neutralizing their energy requirements. While, as discussed earlier, it is impossible to guarantee if the energy they are using is green or “brown,” Google can measure if they are matching their energy consumption with energy from renewable sources across the globe. Such measures are defined as the metric of Power Usage Effectiveness (PUE), an industry metric that will be further explored.
Renewable Energy Credits
In addition to PPAs, another common term Google uses in their documents is Renewable Energy Credits (RECs). In a 2013 document, Google representatives offered a definition of RECs: “A common way that companies seek to support renewable generation is through the purchase of RECs. RECs are a tradable commodity that represent a claim to the environmental benefits associated with renewable power generation – they are not tied to the physical delivery of electrons. RECs are sold either “bundled” with the underlying electricity generated, or ‘unbundled’ as a separate commodity from the energy itself” (Demasi, 2013, p. 2). After Google purchases a renewable energy project contract, they sell that acquired energy to the local energy grid at a “wholesale” price (Google, 2013, p. 3). Once this transaction occurs, Google acquires a “net-loss” since wholesale selling is often less than the renewable acquisition costs. At this point, however, Google states they “strip” the RECs and then “keep them so no one else can claim credit for the green aspect of our purchase” (Google, 2013, p. 3). Another Google document explains that REC purchases are “issued by the renewables industry to record every unit of energy that’s produced by renewable means” (Google Sustainability, n.d., para. 8). With such purchases, Google gets the “credit” for the renewable energy being sold into the power grids. This means that they will “run our facilities with ordinary power purchased from the local utilities and permanently ‘retire’ the RECs against our actual energy consumption, thus reducing our carbon footprint” (Google Sustainability, n.d., para. 8). RECs go beyond PPAs in that Google can sometimes add green energy to local power grids but then get the legal credit for such additions.
Another way Google describes this process includes striping RECs to “keep them so that no one else can claim credit for the green aspect of our purchase” (Google, 2013, p.3). Another document elaborated, stating that “we strip off the newly created RECs from our PPAs (in step 1) and match them to the retail electricity that we purchase at the data center. Over a year, the total number of RECs we apply equals the total consumption at our data center” (Google, 2016, p. 7). So, while PPAs are concerned with adding green electrons from renewable energy projects into the power grid, RECs are the next step in the process, allowing companies to get credit for the amount of renewable energy generated at a project and for no one else to get that credit. Then, since Google knows how much energy it requires to operate, they can know how many PPAs and RECs they need to ensure they can operate neutrally.
Green Tariffs
In addition to PPAs and RECs, another aspect of Google’s energy portfolio includes green tariffs. Green tariffs, according to Celenia Benguli (2019) of the World Resources Institute, are used by companies like Google because “such customers want more than just the Renewable Energy Certifications (RECs) that allow them to claim credibly that they are using green power—they also want access to the long-term, fixedprice structure of renewable energy” (para. 2). Benguli (2019) continues by explaining how Green tariffs “cater to customers’ preference for a more direct financial connection to nearby renewable energy projects” and that “they can also offer greater economic value to customers than unbundled RECs alone” (para. 2). Green tariffs allow energy companies like Duke Energy or Georgia Power to offer customers access to energy from renewable projects located in that region. In regions where energy is regulated, meaning customers have only one choice of company to use to provide their power, green tariffs offer the opportunity for these customers to access renewable energy. Benguli (2019) further explained:
Green tariffs, or riders, emerged as an option for customers in traditional, regulated markets, and have expanded rapidly in recent years. Offered by local utilities and approved by state public utility commissions (PUCs), these programs allow eligible customers to buy both the energy from a renewable energy project and the Renewable Energy Certificates (RECs) at a more favorable price. Green tariffs were originally designed for large-scale energy customers but may include small customers as well. (Benguli, 2019, p. 3)
With all of this in mind, Google’s documents explained green tariffs as “utilities [which] would offer companies like Google the choice to buy renewable energy through a new class of service” however, this voluntary service would only be available to those to “meet basic criteria” (Benguli, 2019, p. 3). If Google spends the money to procure the renewable energy, any customer who then wishes to access the energy would have to pay for the option, which avoids the “impact on other ratepayers” (Demasi, 2013, p. 3). In other words, for those who do not wish to have access to renewables, they would not have to pay for the added projects on the grid they use. For those who want to opt-in, the 2013 Google report lists out a “proposed structure” that includes definitions of “eligibility,” “integrated service,” “renewable generation sources,” “green attributes,” and “pricing” (Demasi, 2013, p. 4). So, while they say that “the concept of a renewable energy tariff” is simple,” Google’s structure reveals the complexity of the process (Demasi, 2013, p. 3).
It is through these renewable options and relationships that Gary Demasi, Google’s Director of Global Infrastructure, discloses their hopes:
We’ll continue to find creative ways to supply our facilities with renewable energy, but we think this solution can provide an important new way to increase the use of renewable energy nationwide. We look forward to working with utilities, state utility commissions, companies and other stakeholders to make it a reality. (Demasi, 2013, para. 5).
By 2018, Demasi echoed similar language about partnerships with utilities companies, concluding with announcing a “public pledge to triple our renewable energy purchases for our data centers by 2025” because “we know we have a lot more work to do” (para. 5). Also in 2018, Michael Terrell, Google’s Head of Energy Market Development, provided an example of how Google is accomplishing their solutions, citing a recent partnership between their Lenoir, North Carolina data center and “local electricity supplier” to establish one of the first utility solar purchase programs in the U.S. (Terrell, 2018, para. 4). Google continually positions itself as a leader in renewable energy, as uniquely positioned for efficiency, and as a developer of partnerships that will be transformative across the world.
“Additionality”
Google, not surprisingly, also discloses how they are strategic when considering which renewable energy purchases they make, beyond ensuring they are located on the same energy grids as their facilities. In 2019, Google’s CEO Sundar Pichai wrote that the most recent renewable energy purchases they made both increased Google’s energy portfolio by more than 40%, but the purchases also met the “rigorous ‘additionality’ criteria we set out long ago for our energy purchases” (Pichai, 2019). As part of their corporate strategy to “ensure that Google is the driver for bringing new clean energy onto the grid,” Google requires any renewable energy projects they purchase to be “additional” (Google, 2016, p. 6). Rather than buying renewable energy from projects that are already inputting power into the local grid, Google mandates their purchases must come by adding brand new projects. Specifically, they state that “we seek to purchase energy from not yet constructed generation facilities that will be built above and beyond what’s required by existing energy regulations (like state renewable energy standards)” (Google, 2016, p. 6). Ultimately, through this requirement of additionality, Google wants to further the economic growth and technological advancements of the regions in which they operate (2016, p. 6). In 2019, this additionality standard reportedly spurred “the construction of more than $2 billion in new energy infrastructure, including millions of solar panels and hundreds of wind turbines spread across three continents” and “our renewable energy fleet now stands at 52 projects, driving more than $7 billion in new construction and thousands of related jobs” (Pichai, 2019, para. 4). These documents explaining additionality again reiterate Google’s stance as an industry leader and economic driver in economies across the globe, spanning beyond the United States.
Measuring Energy Usage- PUE
Beyond matching their energy consumption, Google is concerned with ensuring their internal operations are energy efficient to reduce consumption needs. To do so, Google documents indicate PUE as a metric that involves measuring energy usage “every 30 seconds” in a standardized way across the industry (Kava, 2014). For Google, this means they benchmark their own PUE of their facilities against other competitors, calculating their performance and needs and contrasting their progress. Globally, cloud computing involves data servers, cooling infrastructures, and the warehouses and materials that enable connection on Google’s properties. Energy is required for the functionality of all of these components and efficiency is important to Google, and Google is sure to include “all sources of overhead in our efficiency metric” (Google Data Centers, n.d.-a, para. 8).
The process of measuring energy usage across Google operations is not dissimilar to the energy companies coming to customers’ homes and viewing the meter numbers. In fact, Google describes getting their PUE measurements by using “multiple on-line power meters in our data centers to measure power consumption over time” (Google Data Centers, n.d.-a, para.11). Google notes that “the data center industry uses the measurement ‘PUE,’ or power usage effectiveness, to calculate the energy costs of housing and cooling servers” and that “PUE measures how much overhead energy is required to house and cool the computers inside a building relative to the amount the computers consume themselves” (Google, 2011, p.3). Additionally, Google attempts to ensure they use a more holistic standard than often used in the industry by reporting a more “comprehensive” year-long PUE in all seasons with “all sources of overhead”included in the measure (Google Data Centers, n.d.-a, para. 8). In the document, Google gives access to their comprehensive PUE reports across the years, including the scores from each of their data centers around the globe, spanning from 2008 to 2019 (Google Data Centers, n.d.-a). Though the PUE calculations are advanced mathematical equations, it is worth noting that Google touted an average 2019 PUE score across all data centers as 1.11, “making our data centers among the most efficient in the world” (Google Data Centers, n.d.-a, para. 9). Further in the document, Google disclosed that the largest industry respondents to the Uptime Institute’s 2019 Data Center Survey had a PUE score of approximately 1.67 (Google Data Centers, n.d.-a). The lower the PUE score, the more efficient the operation.
To explain how Google manages to have a lower PUE score, the company discloses how they have managed to find ways to utilize their infrastructure and systems more consciously. One example they report is raising the temperature of their data centers to “80°F,” also “using outside air for cooling” and building their own “custom servers” (Google Data Centers, n.d.-a, para. 5). In doing this, Google can capitalize on some of the regional temperatures for their data centers, as well as perfecting efficiency within their machines by engineering them internally. Additionally, a Google report offers figures that compare their cloud against the in-house servers that companies use, making a case for why they are the most efficient cloud computing option.
Fig. 1: Google’s Energy Efficient Cloud, (Google, 2012, p. 2)
In Fig.1, Google’s 2012 report explains how their cloud solution is more energy efficient versus those using private, in-house data centers. In Option 1, Google highlights how in-house data servers are often more numerous, the cooling of the servers as being “inefficient” and more numerous in terms of equipment to power the “significant network traffic” (Google, 2012, p. 2). Contrast this to Option 2, the image demonstrates how Google has mastered energy efficiency of their data servers, the cooling required, and how ultimately despite a small increase in the amount of energy needed to meet the server network traffic demand, their cloud is the best option (Google, 2012, p. 2). Google’s (2012) report concludes with a staggering statistic: “According to a recent study by the Carbon Disclosure Project, by migrating to cloud computing, large U.S. companies could achieve annual energy savings of $12.3 billion and carbon reductions of 85.7 million metric tonnes by 2020—equivalent to the annual emissions of over 16.8 million passenger vehicles” (Google, 2012, p. 4). Since Google simultaneously positions themselves as an industry leader in operations efficiency, transitions to their cloud services would be the most effective, explicitly stating so in the following: “Because of our energy efficiency efforts, our cloud is better for the environment. This means businesses that use our cloud-based products are greener too” (Google Data Centers, n.d.- a, para. 4).
Another Google report argues that Google’s cloud services capitalize on the scale of their operation, arguing that “cloud providers take advantage of this efficiency in scale by providing servers for millions of users—maximizing the utilization of machines while cutting down on the total number of servers required. The result is fewer machines and less energy over all” (Google, 2011, p. 4). Google’s size and scale of global operations means they can maximize to meet the demands of millions of consumers while being more efficient through strategic engineering. Thus, efficiency means reducing the amount of energy needed to complete a search, streaming a video, sending an email, etc. For example, Google explains how checking an email consumes energy from the client’s device, the network of wireless routers, and the server that sends and stores the emails (Google, 2011, p. 1). A “cloud-based email system” is described by Google as “more efficient” because it “saves considerable amounts of per-user energy costs once provisioning email servers, providing redundancy, and cooling costs are taken into account” (Google, 2011, p. 1). Again, Google reiterates how the scale of their operations offers more efficient options for users.
Google takes their PUE seriously while explicitly promoting how much more efficient it is for other businesses to use their data centers for their cloud storage needs. In the same document that explains their PUE, Google includes a study that “has shown that businesses that use Gmail have decreased the environmental impact of their email service by up to 98% compared to those that run email on local servers” (Google Data Centers, n.d.-a, para. 3). Since Google has mastered their own efficiency, they, by default, also explicitly position themselves as the answer for efficiency for other companies. For example, a 2012 Google report gave an example of the U. S. General Services Association (USGSA) using their services. The U.S. General Services Association is a government organization which contracts, leases, and manages government buildings, as well as offering private sector IT, supplies and equipment, and services to the military (USGSA, 2020). Additionally, the USGSA touts itself as promoting “management best practices and efficient government operations through the development of governmentwide policies” (USGSA, 2020). In 2012, a Google PUE report used the USGSA as a case example, stating that its approximately 17,000 users switched to using Google’s cloud services and saved “$285,000 annually” in energy costs, while also reducing “server energy consumption by nearly 90%,” and “carbon emissions by 85%” (Google, 2012, p. 1). The inclusion of this government organization is important because, as Google makes the claim that they offer the most energy efficient business operations among their competitors, they also give an example of an organization saving with their services that promotes itself as efficient in operations across the government (USGSA, 2020).
Artificial Intelligence/Machine Learning for Efficiency
Google can generate their PUE metrics because of their ability to track their usage data and then optimize their processes. In 2014, an official blog post revealed that one of Google’s data center engineers, Jim Gao, used machine learning to create a system in which a computer could learn from large amounts of data to identify patterns and then “learn” from them to optimize energy usage (Kava, 2014). Gao’s report revealed that “Machine learning is well suited for the DC environment given the complexity of plant operations and the abundance of existing monitoring data” and that “The interactions between these systems and various feedback loops make it difficult to accurately predict DC efficiency using traditional engineering formulas” (Gao, 2014, p. 2). To overcome such challenges, Gao employed “neural networks;” these are complex mathematical equations which access multiple data points simultaneously and configure the rates, in this case, of optimum energy efficiency across operations (Gao, 2014, p. 3). Ultimately, Google can use their machine learning to leverage “the plethora of existing sensor data to develop a mathematical model that understands the relationships between operational parameters and the holistic energy efficiency” in attempts to ultimately reduce their overall PUE across operations (Gao, 2014, p. 7). Gao’s machine learning discoveries have again positioned Google as a leader in energy efficiency for their data centers, promising future efficiency at PUE levels well below the industry standards.
In addition to disclosing information through scientific reports that are linked from Google’s blog, the company also released documents that are more user-friendly to further explain how they are using machine learning to become more energy efficient. On their “Environment Projects” tab through their sustainability website, Google explained how “Machine learning gives computers the ability to learn things without being explicitly programmed, by teaching themselves through repetition how to interpret large amounts of data” (Google, n.d. -d, para. 5). Google also reiterates how such machine learning is already used by the company in search engine capacities: “Google already uses it to improve features like translation and image recognition. When you ask Google Photos for pictures of people hugging, it’s machine learning that finds the photos you’re after” (Google, n.d. -d, para. 5). Because machine learning can analyze so many different, complex data points across the enterprise simultaneously and then configure options for more efficiency, Google describes the future potential: “Google’s environmental team wants our operations to emit less carbon. Hardware ops aspires to fewer component failures. The platforms people care about server energy consumption. Machine learning can help them all achieve their efficiency dreams” (Google, n.d. -d, para. 10).
As an industry leader in the renewable energy acquisition and PUE metrics, Google also positions itself as a leader in artificial intelligence and machine learning in other documents, stating that they are “using AI to build a more sustainable world” (Porat, 2019, para. 15). Ruth Porat, Google’s SVP and CFO, wrote that “We built an AIpowered efficiency recommendation system that directly controls data center cooling” and “this first-of-its-kind cloud-based system is delivering energy savings of roughly 30 percent” (Porat, 2019, para. 16). Porat further asserted that, “After DeepMind and Google started applying machine learning algorithms to 700 megawatts of wind power in the central U.S., the value of that wind energy has been boosted by roughly 20 percent”(Porat, 2019, para. 17). Beyond added benefit to wind energy and AI-powered efficiency and in similar themes to creating greener options for all, Gao also argues that their machine learning can “help other companies and industries get a lot greener, in both senses of the word” and plan to release another paper with more details about their developments in the near future (Google, n.d. -d, para. 11). In a collaborative report in 2013 referenced on Google documents, the Lawrence Berkeley National Laboratory research team investigated data centers, and they found that “cloud computing holds great potential to reduce data center energy demand moving forward, due to both large reductions in total servers through consolidation and large increases in facility efficiencies compared to traditional local data centers” (Masanet et al., 2013, p. 1). Through their partnerships with outside research teams and their own internal developments with artificial intelligence to improve their PUE and overall efficiency, Google positions itself as an incredibly viable option for customers and a leader over competitors.
Google’s Circular Economy
My analysis also revealed mentions of “a circular Google” and the “circular economy” while positioning itself within the climate crisis (Brandt, 2019, para. 3). In 2019, Kate Brandt, Google’s Chief Sustainability Officer, published an article on the “Outreach & Initiatives” section of Google’s website, introducing the “circular economy” model and outlining why it is an essential part of Google’s current and future practices as a company (Brandt, 2019, para. 3). Brandt (2019) begins by introducing how the consumer economy “demands” more than can reasonably be produced by the Earth- “just last year, humanity’s consumption of resources– such as metals, timber, and even land– required 1.7 planet Earths to sustain”(para. 1). After another paragraph dedicated to explaining waste from plastic straws to plastic dumped into the oceans, Brandt explains how “for too long, the damaging environmental consequences of these linear systems remained relatively invisible,” but now, “the impact cannot be ignored” (Brandt, 2019, para. 2). After discussing the current issues related to resource consumption, the post continues by introducing Google’s new “circular strategy” that “is part of our wider effort to build sustainability into everything we do” (Brandt, 2019, para. 4).
Another document argued that “today’s economy is linear: it has a beginning and an end” and this means that “companies dig up materials, turn those materials into a product, and then ship that product to an end user who eventually tosses it in the trash” (Google, n.d. -c, para. 3). In light of the climate crisis, Google promotes an understanding of how such linearity is not only damaging, but the “system has to change” (Google, n.d.- c, para. 3). Focusing on their data center operations specifically, Google partnered with the Ellen MacArthur Foundation in 2016 to perform a case study of data centers and the circular economy model (Brandt & Rana, 2016). Google described choosing the partnership with the Ellen MacArthur Foundation because it is “a nonprofit that helps companies around the world adopt circular economy practices and experience the enormous benefits” (Google, n.d.-c, para. 5). Google also explained they chose to do a case study of their data centers specifically “because data centers generally tend to be material intensive,” and “they are like small cities filled with servers, drives, routers and other components” (Google, n.d. -c, para. 6).
The 2016 case study revealed Google’s three principles for their circular economy: “preserve and enhance natural capital by controlling finite stocks and balancing renewable resource flows,” “optimise resource yields by circulating products, components, and materials,” and “foster system effectiveness” through design (Brandt & Rana, 2016, p. 7). As illustrated in Fig. 2 below, the circular economy model has four elements that include maintenance/prolonging, refurbishing/remanufacturing, reusing/redistributing, and recycling (Brandt, 2019). The 2016 case study explained each of these elements. The maintenance step involves data center repairs programs that “replaces failed components using a mix of new and refurbished parts” (Brandt & Rana, 2016, p. 4). The refurbish step involves the custom building of their own servers, evaluating which parts can be refurbished (Brandt & Rana, 2016, p. 4). The reuse step involves reselling components to “selected remarketing partners,” and in 2015 alone, Google “resold nearly 2 million units into the secondary market for reuse by other organizations” (Brandt & Rana, 2016, p. 5). Finally, the recycle step involves crushing and shredding the “electronic equipment” like storage tapes and hard drives that cannot be resold (Brandt & Rana, 2016, p. 6). After the crushing and shredding occurs, the remains are “sent to a recycling partner for secure processing” and creation of reusable materials (Brandt & Rana, 2016, p. 6). In 2019, Brandt explained how implementing this circular economic model “could generate $4.5 trillion in new economic output by 2030,” further expressing how “abundance” and “progress” are not mutually exclusive from “improving human and environmental systems” (Brandt, 2019, para. 7).
Fig. 2: Google’s Model of a Circular Economy (Brandt, 2019)
Though much of the analysis in this thesis oscillates around energy consumption and Google’s renewable energy acquisitions and requirements, it is worth including this section on the circular economy because again, it contextualizes the choices and operations Google makes and expresses how they understand their corporate responsibility. The first principle of the circular economy model explicates renewable resources (Brandt & Rana, 2016, p. 7). Additionally, through such careful explanation of their circular economy processes, Google can again position themselves as a leader over competitors in keeping their core commitments to sustainability in everything they do. In fact, the 2016 case study concluded with “it is Google’s belief that doing so will yield additional value for business, partners and users” (Brant & Rana, 2016, p. 8). Brandt reinforced the circular economy as something essential: “our goal is to embed circular economic principles into the fabric of Google’s infrastructure, operations, and culture” (Google, n.d. -c, para. 15). A circular economy is positioned as the most reasonable move, both sustainably and financially, for a corporation of Google’s caliber. In 2019, Brandt stated that the information they collect and data they gather within their facilities informs how to be more efficient in operations is useful so “we can leverage our scale, resources, and technological expertise to help the world meet resources needs” (Brandt, 2019, para. 6). Again, the circular economy appears to be one way Google explicates their role as a corporate leader that is the most environmentally sustainable option for consumers.
This chapter provided an overview of the key emerging themes throughout Google’s public-facing documents related to their environmental sustainability, their requirements for renewable energy acquisitions, their vast renewable portfolio, and the various company initiatives like the “circular economy” (Brandt, 2019). Such findings were important because they demonstrate the discourse the company has used and is currently using as they address their role in the climate crisis for the public. These findings also reveal the common terms Google uses while describing their efforts in sustainability and how they understand themselves as a benevolent, corporate answer. Having done this work, I now turn to an analysis of this material foregrounding of what Google hopes to gain or accomplish by positioning themselves in this way.