Public Comment: T&E Committee Meeting 02/09/26,
Item 1 - Large-Load Energy Customer Development
Status Report
I founded AllAI Consulting, LLC to non techies, specifically lawyers, advocates, and public officials. It was in that capacity that I provided both oral and written testimony for the Little Hoover Commission hearing that was referenced in the Memorandum. As part of this education effort, AllAI Consulting is engaged on a multi-year effort with the Investigations Lab at the UC Berkeley Human Rights Center to map out the regulatory process for getting a data center approved in the State of California. That work requires us to dive deep into the regulatory filings and analysis performed through the various state and local regulatory approval processes (CEQA, Water filings, SPPE’s, and EIR’s). The key takeaway that became apparent very early in this research process is that local governments cannot assume the sufficiency of these external analyses. While they are designed to consider the local context and provide reliable findings that can be wholesale inserted into local government decision-making processes, they often fail to meet that goal for a variety of reasons. Specifically, they include glaring unaddressed contradictions and omissions, outdated and unrealistic approval thresholds, and fail to consider the comprehensive impacts of the entirety of the San Jose data center build out.
These concerns are tangibly demonstrated through a review of the Environmental Impact Review performed on the recently-opened Great Oaks South data center.
Great Oaks South is an Equinix co-location data center located off Great Oaks Boulevard in San Jose.1 As a co-location center, it will host loads from various data center customers in a landlord-tenant style relationship.2 This data center is the first facility delivered by PG&E3 under the Implementation Agreement with the City of San Jose. The facility was required to undergo a formal Environmental Impact Report (EIR), process as part of its application for a Small Power Plant Exemption from the California Energy Commission (CEC). The EIR is the core environmental review performed on a project in California. It’s findings are often ingested whole and unquestioned into city and county-level decisionmaking processes. In reviewing the entirety of the record related to the final approval of the EIR by the CEC4, several issues of local import become clear.
Water - unaddressed contradictions and omissions
Hydrology and Water Quality Finding - The project’s proposed use of 4 acre-feet (AF) of water during construction and 4 acre-feet per year (AFY) during operation would not substantially decrease critical groundwater supplies. The project’s impact on groundwater supplies, recharge, 4 California Energy Commission. Docket: 20-SPPE-01 Great Oaks South Backup Generating Facility Small Power Plant Exemption
3 Stephanie Magallon. PG&E Delivers First San Jose Data Center as Part of Historic Agreement. PG&E. (January 22, 2026)
2 Megan Baggiony-Taylor. Equinix: How Interconnection Drives this Data Centre Giant. (October 14, 2025) 1 Equinix. SV5 Silicon Valley IBX® Data Center
Submitted by Masheika Allgood, AI Ethicist and Founder of AllAI Consulting, LLC or sustainable groundwater management during construction and operation would therefore be less than significant.5
Data centers consume water in two distinct ways, we refer to these different types of consumption as scopes. Scope 1 water consumption is water that is directly consumed by the data center. This is overwhelmingly drinking water and it is consumed while cooling the data center equipment. Scope 2 water consumption is water that is indirectly consumed by the data center. This is largely a mix of surface and groundwater that is consumed while generating power for the data center. Both types of consumption must be considered in any evaluation of the impact of a data center will have on our finite, and often limited water resources. The EIR process only considers Scope 1 consumption. And in the Great Oaks EIR, a blatant contradiction in the type of cooling system the facility is using caused the lead agency to completely forego any form of Scope 1 consumption evaluation.
Scope 1 - Air cooled v. Water cooled
In the facility Cooling System is described in the Project Description as follows: “Each building would be cooled by an Air-Cooled Chilled Water System with refrigerantside economizer. The new mechanical system would consist of 72 total 400-ton chillers, 24 per building.”6 Based on this description, the Water Use assessment concludes: “The theoretical maximum operational usage for each building is approximately 1.2 acre-feet per year.”7 So, the facility will utilize roughly 1.2 million gallons of water per year using Air-Cooled Chilled Water Systems with air-cooled economizers. However, Table 3-2 outlines the mechanical and electrical systems that the facility will use to optimize energy performance. And, according to that Table, the facility will be using Water-Cooled Chilled Water Systems with water-side economizers8 for the data halls:
8 Cooling Water Efficiency Opportunities for Federal Data Centers: Use of Water-Side Economizing Strategies
7 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-8. Page 64
6 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-8. Page 64
5 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report.
Summary 1-7. Page 12
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 2 Why does this matter? Because there are only two ways to cool a data center - electricity (air conditioning) or water (evaporation). The basic principle is simple. When you finish cooking and want to cool off a hot pan, you run it under water in the sink. The heat from the pan evaporates the water and the pan cools down. You could also hold that pan in front of an air conditioner, but you’d need a freezer-level air conditioner to match the performance of your sink. These AI data center facilities are trying to cool thousands of chips that each run at 176F. Way hotter than your pan will ever get on the stove. So the amount of water that is evaporated during the heat exchange process is massive. And the process is always running since AI data centers never turn off.
It would take a massive energy expenditure to continuously air cool a hyperscale AI data center campus, but if you’re willing to spend the electricity, you can reasonably run an air cooled system on 3.6 acre-feet of water per year. But a water cooled system can consume that much water every single day. The industry standard calculation for determining the evaporation rate of a water-cooled system is:9
9 Jason Chesley. Data Center Water Treatment Systems In Theory and in Practice: Water Uses in AI Data Centers. Ecologix. (October 22, 2025)
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 3 For a 50 MW heat load at 80% efficiency, the rate of evaporation is 26,295 gallons per hour.10 Which translates to over 230 million gallons of water per year. That’s nearly 707 acre-feet per year. Which is right in line with the 1,036.78 acre-feet per year that was requested by the developer11 when the maximum campus electricity load was projected to be 57 MW (19MW per building).12 All of which was to be taken from our groundwater aquifers: “GOWC will supply all potable water for the Project, as well as its present and future customers from groundwater resources.”13
And although the closest recycled water line is located approximately one mile from campus, the facility operator rejected any use of recycled water.14 It was only after being pressed about the “unreasonable” and “wasteful” use of 1,000 acre-feet of potable water for cooling purposes when “recycled water from the South Bay Water Recycling Program (SBWRP) is available in the project area,” that the facility operator responded that they would replace the Water-Cooled Chilled Water system with Air-Cooled Chilled Water System with refrigerant-side economizer specifically to reduce the total water demand “to less than 4 acre feet per year for all three data center buildings.”15
But the EIR illustrates the facility operator trying to have it both ways. When pressed on the unreasonable amount of potable water requested for cooling operations, they claim to move to electricity-driven cooling systems. But when asked to list measures to optimize energy performance, they provide a clear preference for water cooling. In Table 3-2, within the HVAC measures, the facility operators indicate that they will actively attempt to increase the amount of time the water-side economizer can run - “with A/C compressors operating only during peak load periods.”16
Water and electricity have an inverse relationship in data center cooling. The facility operator attempts to address this inverse relationship by describing the facility as water cooled where the agency is concerned about electricity, and air cooled where the agency is concerned about 16 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-15. Page 71
15 TN233005-4 - SV1 Responses to CEC Data Requests Set 2 GOSBGF - Part IV. Page 5 14 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-8. Page 64
13 TN233005-4 - SV1 Responses to CEC Data Requests Set 2 GOSBGF - Part IV. Page 9 12 TN232467-3 - GOSBGF SPPE Application Appendices H-K. Page 137 11 TN233005-4 - SV1 Responses to CEC Data Requests Set 2 GOSBGF - Part IV. Page 13 10 Jason Chesley. Data Center Water Treatment Systems In Theory and in Practice: Water Uses in AI Data Centers. Ecologix. (October 22, 2025)
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 4 water. Whether the facility is actually air cooled, water cooled, or a hybrid of both is unclear from the Final EIR. And there’s currently no one within City government who’s tasked with finding out.
Wastewater Treatment - cooling system blowdown unaddressed When the applicant submitted the SPPE application, the project design included the use of a water-cooled chilled water system instead of the now-proposed air-cooled system.
The applicant did not provide information about the revised wastewater amounts the project would generate as a result of switching the cooling system… The RWF has a capacity to treat 167 million gallons per day (mgd) of wastewater and currently treats an average of 110 mgd, leaving RWF with 57 mgd of available capacity, which is substantially more than the project’s discharge of approximately 650 gpd. Therefore, there is an abundance of capacity at the RWF to accommodate project flows.
17 Data centers utilize a variety of physical and chemical filtration processes to ensure the water in their cooling systems doesn't negatively impact the equipment. The pre-treatment process begins when water is initially ingested into the facility. The water is filtered through screens and then undergoes reverse osmosis to remove impurities that could cause scaling or corrosion.
Then softeners or ion exchange systems are used to reduce water hardness. And antiscalants (e.g., phosphonates) and pH adjusters are used to minimize scaling and fine tune water hardness levels.18
Once the water enters the cooling system, it accumulates chemical additives like anticorrosives (e.g., molybdate), biocides, and leached metals (Cu, Zn).19 And the blowdown from the cooling towers, the water that is drained from the system to keep the pollution levels stable,20 can concentrate total dissolved solids in the water to 2,000 ppm.21 A cooling system’s blowdown is calculated as:22
COC, or cycles of concentration, refer to the amount of freshwater you put into the cooling system to offset the amount of water that has to be drained due to pollution. Data center COC values range from 4 to 6, where a value of 4 is more blowdown intensive and a value of 6 is more water intensive. But the amount of blowdown that a data center produces is significant. If 22 Jason Chesley. Data Center Water Treatment Systems In Theory and in Practice: Water Uses in AI Data Centers. Ecologix. (October 22, 2025)
21 Jason Chesley. Data Center Water Treatment Systems In Theory and in Practice: Water Contamination Mechanisms. Ecologix. (October 22, 2025)
20 Cooling Water Efficiency Opportunities for Federal Data Centers: Maximizing cycles of Concentration 19 Jason Chesley. Data Center Water Treatment Systems In Theory and in Practice: Water Contamination Mechanisms. Ecologix. (October 22, 2025)
18 Jason Chesley. Data Center Water Treatment Systems In Theory and in Practice: Sources and Pre-Treatment of Water. Ecologix. (October 22, 2025)
17 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Utilities and Service Systems. 4.18-9. Page 373
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 5 we use the example from the Scope 1 section, where the system evaporated 230 million gallons of water per year, the blowdown for a system with 4 COC would be over 76 million gallons of water per year. The blowdown for a 6 COC system would be over 38 million gallons of water per year.
This blowdown will need to be treated before it is released into communal water systems.
However, given that Hydrology and Water Quality finding is based on an air cooling systems, the EIR does not discuss the burden water cooled systems would put on our wastewater treatment facilities. The evaluations in the Utilities and Service Systems section of the Environmental Requirements for Data Center Development in San Jose Attachment focus on the capacity of our wastewater treatment systems, not whether the infrastructure can manage the flow rates required for direct receipt of data center wastewater, or if the current wastewater treatment process23 is well suited for the specific mix of concentrated pollutants that data center blowdown contains. And there’s currently no one tasked at the city, county, or state level with assessing the burden data centers put on our wastewater treatment infrastructure from an equipment or methodology perspective.
Scope 2 - Indirect water consumption assessment
A data center’s Scope 2 water use often dwarfs the Scope 1 use.24 As part of my written testimony before the Little Hoover Commission, I derived the amount of water that is withdrawn to produce 1kWh of electricity on California’s grid. Currently, 5.2 gallons of water is withdrawn to produce 1 kWh of electricity in California, with 10% of that water lost to evaporation.25 The EIR “conservatively assumes that the [Great Oaks facility] could consume up to 867,240MWh per year, based on the maximum demand of 99 MW during all 8,760 hours per year.”26 Given California’s grid composition, generating that much electricity will require the withdrawal of 4.5 billion gallons of water, with over 450 million of those gallons lost to evaporation.
The EIR doesn’t include any discussion or assessment of the campus’ Scope 2 water usage.
Neither do any of the processes in the Water Management section of the Environmental Requirements for Data Center Development in San Jose Attachment. The U.S. Energy Information Administration is starting to make energy generation facility-level water consumption data public, but the feature is only in beta at the moment.27 And there’s currently no one tasked at the city, county, or state level with assessing the impact of Scope 2 data center water consumption on our local or state water resources.
27 EIA Electricity Data Browser - Beta
26 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report.
Greenhouse Gas Emissions 4.8-10. Page 269
25 Masheika Allgood. Little Hoover Commission Written Testimony. (November 17, 2025). Page 3 24 Luke Barratt and Rosa Furneaux. Amazon strategised about keeping its datacentres’ full water use secret, leaked document shows. The Guardian. (October 25, 2025) 23 San José-Santa Clara Regional Wastewater Facility: Treatment Process Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 6 Air Quality
Air Quality Finding - the project would not cause a cumulatively considerable net increase of any criteria pollutant, and impacts would be reduced to less than significant with mitigation incorporated.
28 AI Data Centers are increasingly considered critical infrastructure. As such, they are required to run at high availability, which is also referred to as over 99% uptime.29 The uptime values translate to actual time values, as indicated in the following chart:30 A facility that commits to 99.999% uptime (“five 9’s availability”), is only allowed 5 minutes of downtime per year. Which means that the facility has an extreme dependence on unquestionably reliable backup power. Diesel generators are historically reliable backup generation sources and, given the nearly non-existent tolerance for failure for high availability facilities, data center operators are justifiably cautious about implementing any backup system technology that does not have an equally long and verifiable track record. The Great Oaks South Facility will operate an installation of 36, 3.25 MW diesel-fired generators.31 31 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-7. Page 63
30 Joe Hertvik. What Is Five 9s in Availability Metrics? Splunk. (August 16, 2024) 29 Joe Hertvik. What Is Five 9s in Availability Metrics? Splunk. (August 16, 2024) 28 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report.
Summary 1-4. Page 9
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 7 Modeling Air Quality Health Impacts - local and state expert analysis overruled
Diesel generators aren’t infallible. Facilities must run their installations for hours every month to make sure they're working properly. The amount of time these generators run is of paramount importance to any agency charged with the health and welfare of neighboring residents.
According to the EIR, the facility operator proposed “an annual readiness and maintenance testing schedule that would be comprised of 20 hours per year (per engine).”32 The EIR also asserts that the facility’s generators “would be run for short periods for testing and maintenance purposes and otherwise would not operate unless there is a disturbance or interruption of the utility supply.”33
This point of fact was strenuously argued and actively refuted by the Bay Area Air Quality Management District (BAAQMD).34 BAAQMD provided evidence35 of data centers under their purview operating their backup generators “more frequently than previously understood because of climate change induced crises and grid operational challenges.”36 According to BAAQMD:
Between September 1, 2019, and September 30, 2020 nearly half of the identified data centers in Santa Clara, San Jose, and Sunnyvale operated backup diesel generators for reasons other than routine testing and maintenance. Many of the data centers operated diesel generators during multiple nontesting/non-maintenance events; non-testing/ non-maintenance hours of operation approached 50 hours for one generator for one event; it appears 40 or more generators operated concurrently at two facilities; and one facility ran diesel generators for approximately 400 hours for non-testing/ non-maintenance purposes over the course of the period.37 Despite the BAAQMD’s insistence on the need for real-world scenarios to be incorporated into the scenarios for environmental analysis, and the need to evaluate the “potential cumulative health risk impacts of TAC and PM2.5 emissions on sensitive receptors within and near the Project area,”38 neither suggestion was followed in the Final EIR. Concerns about the failure to realistically model the real world impacts were raised by a representative of California Air 38 TN235803 - Bay Area Air Quality Management District Comments - Comment Letter for Great Oaks South Data Center NOP. Page 4
37 TN235803 - Bay Area Air Quality Management District Comments - Comment Letter for Great Oaks South Data Center NOP. Page 4
36 TN238700 - Bay Area Air Quality Management District Comments - Comment Letter for Great Oaks South Backup Generating Facility DEIR. Page 5.
35 TN235803 - Bay Area Air Quality Management District Comments - Comment Letter for Great Oaks South Data Center NOP. Page 7
34 TN238700 - Bay Area Air Quality Management District Comments - Comment Letter for Great Oaks South Backup Generating Facility DEIR
33 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-14. Page 70
32 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Project Description 3-14. Page 70
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 8 Resources Board in a lead agency committee status conference,39 as well as in Robert Sarvey's Post-Hearing Brief40 and his Comments on the Proposed Decision.41 To no avail.
Our local and state air quality agencies had good reason for concern. There is no safe level of PM2.5 pollution and severe adverse health impacts can occur with 24 hours of exposure.42 According to the EPA, the small particles that make up PM2.5 pollution can get deep into your lungs, and your bloodstream. Once these fine particles enter your bloodstream they can attach themselves to your red blood cells43 where they can cause a variety of health impacts including:44
● premature death in people with heart or lung disease
● nonfatal heart attacks
● irregular heartbeat
● aggravated asthma
● decreased lung function
● increased respiratory symptoms, such as irritation of the airways, coughing or difficulty breathing
The lead agency refused to update its model scenarios and even dismissed the facility operator’s self-imposed generator runtime limit. In the initial submission, when the campus intended to run 21 generators, the facility operator proposed to limit generator testing and maintenance to “no more than 356 hours in any consecutive 12-month period” out of concern for NOx emissions, but in the EIR the lead agency found that limitation was not necessary since “BAAQMD’s permitting process will ensure that the project provides all necessary offsets from a location(s) sufficient to mitigate impacts.”45
Under the current laws, California allows modern industrial sized diesel generators to be run 50 to 100 hours a year for maintenance and testing, with some facilities being allowed to run 100 hours a year.46 While older, more polluting models may run for 20 hours a year.47 Since large load facilities are so large, it is critical that they be encouraged to drop off of the grid in times of severe grid stress. In California, that usually means one of our increasingly long and frequent heat waves, where residential and office customers are using atypical amounts of power. Facility operators are allowed up to 50 hours of non-emergency backup power generation for these grid 47 California Air Resources Board. Use of Back-up Engines for Electricity Generation During Public Safety Power Shutoff Events
46 17 Cal. Code Regs. § 93115.6(a)(3)(A)
45 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Air Quality 4.3-24. Page 124
44 EPA. Health and Environmental Effects of Particulate Matter (PM) 43 Lancaster University. Researchers discover air pollution particles hitching a ride around the body on red blood cells. (October 14, 2025)
42 California Air Resources Board. Inhalable Particulate Matter and Health (PM2.5 and PM10) 41 Robert Sarvey's Comments on the Proposed Decision
40 Robert Sarvey’s Post-Hearing Brief
39 Robert Sarvey’s Post-Hearing Brief. Page 6
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 9 stress relieving purposes.48 Facility operators are also allowed to run their backup installations without limitation when the local utility performs a Public Safety Power Shutoff (PSPS).49 But, in practice, a local data center operator ran their diesel generators for approximately 400 hours in a single year. And there’s currently no one tasked at the city, county, or state level with reviewing the actual runtime of local data center generator installations and assessing the cumulative impact of data center backup generators on our local air quality.
Model assumptions - inversion not accounted for
According to the American Lung Association, the Bay Area is the 7th most polluted city in the country in terms of fine particle pollution.50 The Bay Area’ unique geography creates a near-daily inversion layer that serves to trap and concentrate pollutants close to the ground for large portions of the day.51 This inversion layer is a known feature of the Bay Area, but it isn’t included in the scenarios or thresholds for the air quality evaluation in the EIR. In fact, the EIR asserts that “The worst-case impacts of the diesel emergency standby engines would occur at or near the fence line and decrease rapidly with distance from fence line.”52 That is clearly not how an inversion layer works. The air is trapped and gets stagnant,53 and we re-breathe it until the inversion lifts. There’s currently no one tasked at the city, county, or state level with reviewing air quality modeling and assessing whether it accounts for actual local conditions.
Pollution thresholds - outdated
The first filing in the EIR process was on Mar 18, 2020.54 As part of that filing, the facility operator performed an initial air quality evaluation.55 Table 4.3-1 lists the Air Quality Significance Thresholds for that time. According to the chart, the BAAQMD’s PM2.5 thresholds were 54 pounds per day Average Daily Emissions and 10 tons per year Annual Average Emmissions.56 The Final EIR provides an updated version of this chart that lists the State and National thresholds in the now standard unit of micrograms per cubic meter (μg/m3 ).57 In this Table, the threshold for 24-hour Emissions is 35 μg/m3 and the threshold for Annual Mean Emissions is 12μg/m3 for primary and 15 μg/m3 for secondary.58
58 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Air Quality 4.3-6. Page 106
57 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Air Quality 4.3-6. Page 106
56 TN232467-3 - GOSBGF SPPE Application Appendices H-K. Page 99 55 TN232467-3 - GOSBGF SPPE Application Appendices H-K. Page 107 54 California Energy Commission. Docket: 20-SPPE-01 Great Oaks South Backup Generating Facility Small Power Plant Exemption
53 Zoe Mintz. California Central Valley fog, poor Bay Area air quality explained. CBS News. (February 3, 2026)
52 TN239063 - Great Oaks South Backup Generating Facility Final Environmental Impact Report. Air Quality 4.3-13. Page 113
51 Awair. Why California Cities Rank Highest for Air Pollution. (August 20, 2019) 50 American Lung Association. State of the AIR 2024 Report 49 California Air Resources Board. Use of Back-up Engines for Electricity Generation During Public Safety Power Shutoff Events
48 Sidley Austin LLP. EPA Clarifies Rules for Backup Generator Use. (May 15, 2025) Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 10 Particulate matter is one of the six criteria pollutants considered harmful to public health and the environment under the Clean Air Act, so the EPA periodically review air quality standards related to particulate matter to ensure that they provide adequate health and environmental protection.59 The EPA performed its most recent review of PM2.5 standards in 2024 and revised the Annual Mean Emissions standard from 12μg/m3 for primary to 9μg/m3. 60 This revision is
significant due to the level of baseline 2.5PM pollution that exists in the City of San Jose. Table
4.3-9 lists the background PM2.5 in the facility area:
The background PM2.5 pollution in the area was 108% of the EPA standard. Now that the standard is 9μg/m3, the background PM2.5 pollution in the area is 140% of the EPA standard.
There is a real question of whether any new PM2.5 emitting uses should be approved for the area while the baseline pollution remains so far over the EPA guidelines.
The community is already suffering an overwhelming air pollution burden. Without a reasonable modeling of the impact of the generator installation running under real-world scenarios, we are unable to determine how this data center facility will exacerbate that burden. And there’s currently no one tasked at the city, county, or state level with reviewing the comprehensive impact of data centers in the city, ensuring that the newest EPA standards are applied to both new and existing data center projects, and reviewing actual data center emissions to determine their impact on the residents of San Jose and the broader Bay Area.
Conclusion
The Council cannot assume that the mere existence of these reporting systems guarantees the safety of the residents of San Jose. Because it is clear from examining the EIR that state reports often don’t live up to the standard that we all expect. Things slip through the cracks. And when 60 EPA. Timeline of Particulate Matter (PM) National Ambient Air Quality Standards (NAAQS) 59 EPA. Setting and Reviewing Standards to Control Particulate Matter (PM) Pollution.
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 11 things slip through the cracks at the state level, it is the population of the City of San Jose who have to live with the consequences.
The concerns I raise here aren’t speculative. The residents of Boardman Oregon are currently suffering from the adverse health impacts from super-concentrated nitrates in AI data center wastewater discharge.61 Santa Clara County as a whole, and the San Teresa neighborhood specifically are already living with background PM2.5 levels that are 140% of the EPA’s Air Quality Standards. And states across the country are enacting data center moratoriums62 and legislation to protect their residents from data center environmental impacts and utility rate increases.63
The Appendix seems to represent a justification for abdication. City leadership doesn’t need to directly consider the impact of AI data centers because these other agencies and processes will protect our residents. But in reality, the Appendix is a list of disconnected processes managed by a patchwork of agencies. The City Council is where these disparate process come together in one place. It is where the processes for an individual facility can be looked at comprehensively, and where the impact of multiple facilities can be looked at comprehensively.
There is no other body tasked with this responsibility.
We are in the midst of one of the largest infrastructure buildouts in San Jose’s history. This is not the time to abdicate oversight authority, it’s the time to lean into it. San Jose is the home of the industry that coined the phrase ‘data-driven decisionmaking’. It’s time to put those methods in practice in service of the health and well-being of our neighborhoods and community. San Jose has a responsibility to ensure that these data centers meet our local requirements, which means creating an oversight process that empowers city agencies to validate the sufficiency of the health and environmental impact analyses of all state-level environmental reports.
Specifically, the T&E Committee/Council should consider:
● Holding a public Council Study Session on the cumulative impacts all data center development in San José (existing and planned/in development) ● Creating a Data Center Oversight Committee that is tasked with reviewing the sufficiency of the health and environmental impact analyses of all state-level data center environmental reports. The Committee should include local community members, leaders and experts.
● Commission a study on the real world impacts of currently operating data centers that compares the EIR statements to actual operations.
63 Robert L. Mayo. Data Center Regulation in the United States. North Dakota State University. (October 2025)
62 Georgia; Maryland; Oklahoma; New York; Michigan
61 Sean Patrick Cooper. ‘The Precedent Is Flint’: How Oregon’s Data Center Boom Is Supercharging a Water Crisis. Rolling Stone. (November 24, 2025)
Public Comment T&E Committee Meeting 02/09/26, Item 1 - LLC Status Report 12