What Would a 10-Megawatt Data Center Cost Lodi?

What Would a 10-Megawatt Data Center Cost Lodi?

Before any company breaks ground at White Slough, the City would need to invest between $108 million and $221 million. Here’s where that money would go — and what protects residents from absorbing the cost.

LodiEye Analysis — May, 2026

What This Article Covers

LodiEye’s earlier article on the data center question asked what kind of facility Lodi could realistically host. The answer was: roughly 10 megawatts — about the size of the Nautilus operation in Stockton, or NTT’s facility in Sacramento. Big enough to matter to the City budget. Small enough to be plausible.

This follow-up moves from “what’s possible” to “what would it cost, and how long would it take.” If Lodi decided to host a 10-megawatt data center at the White Slough complex on Thornton Road, the City would have to commit between $108 million and $221 million in infrastructure investment over the next five to seven years — before any data center operator turns dirt.

That number is on the order of Lodi’s entire annual general fund budget. It is also separate from — and additional to — the $60 to $150 million it would cost a developer to build the data center itself.

This article walks through where that money would go, what could go wrong, and how Lodi could structure a deal that keeps residents from ending up paying for a facility that’s supposed to benefit them. The numbers are anchored to comparable Central Valley projects in Manteca, Tracy, and Sacramento. The ranges reflect real uncertainty rather than guesswork.

A note on the hydrogen plant. The Lodi hydrogen project lost its $35 million federal funding in October 2025, and that situation remains unresolved as of this writing. This article treats the hydrogen plant as a separate question from the data center, because the data center conversation can move forward (with a weaker marketing pitch) even if the hydrogen project doesn’t. LodiEye’s earlier recommendation still stands: hydrogen plant construction milestones should be met before any data center entitlements are issued.

Why 10 Megawatts Is the Realistic Number

Ten megawatts isn’t a guess. It’s the threshold between hosting a customer and reshaping the whole utility.

Below 5 megawatts, a data center fits inside Lodi Electric Utility’s existing system with modest upgrades — the kind of work the City would probably do anyway for general industrial growth. Above 15 megawatts, the City would have to redesign the utility’s distribution and subtransmission network from scratch — work that wouldn’t happen otherwise, and that residents would end up paying for through their utility bills. Ten megawatts sits at the upper edge of what’s manageable without that kind of overhaul.

Ten megawatts also matches data centers people can actually go see. Nautilus Stockton 1 runs at about 6.5 megawatts. NTT’s Sacramento facility runs at 12.6. Quest’s Roseville facility runs at 8. These aren’t Google or Microsoft hyperscale campuses. They’re large industrial buildings with very high power demand — the size of one or two big warehouses, not a corporate campus.

At that scale, a Lodi facility would bring the City between $2.4 and $4.3 million a year in revenue once operating — meaningful for the budget, but not transformative. Roughly enough to cover the structural budget shortfall Lodi has projected for the next five years, with some left over for capital projects.

To give a sense of scale: a 10-megawatt data center would use about 80 million kilowatt-hours of electricity a year. That’s roughly 18 percent of what Lodi Electric Utility sold to all its customers combined in 2023. It would use between 30,000 and 200,000 gallons of cooling water a day, depending on what cooling system the operator chose. It would occupy 30 to 50 of the roughly 90 acres available at White Slough after PG&E’s planned hydrogen research facility takes its 130-acre share. And it would create 25 to 35 permanent jobs paying $80,000 to $140,000, plus 200 to 300 construction jobs during the two-to-two-and-a-half-year build.

What Lodi Would Have to Build First

Before any data center operator can turn dirt at White Slough, the City of Lodi, Northern California Power Agency, and Pacific Gas & Electric have to install the infrastructure that makes the site usable. This is separate from the data center building itself — it’s all the work that has to happen on the way to the site and at the site so the building can be built and operated.

Six pieces make up this work. A seventh — the hydrogen plant funding gap — is treated as a separate question because it’s primarily NCPA’s exposure, not the City’s, and because the data center conversation can move forward independent of it (with a weaker marketing story).

Where the Money Goes

Electric infrastructure

$30 – $60 million

A new electrical substation — the size of one of the larger buildings on Industrial Way — to step down high-voltage power and feed it to the facility. Plus the medium-voltage lines connecting that substation to either NCPA’s system or PG&E’s regional grid. The catch: transformers right now have a 2-to-4-year wait time worldwide. LEU Director Jeff Berkheimer told Council on May 6 that the electric distribution system to serve a data center at White Slough isn’t built yet.

Fiber and network

$10 – $20 million

Pull fiber cable from I-5 into White Slough — about a kilometer and a third, similar to running a line from Industrial Way to the WPCF. Build a backup path through a separate route, modeled on the 18-mile fiber ring Utility Telecom built for the Nautilus data center in Stockton. Add a small building where different fiber carriers connect to each other. The closest planned State Middle-Mile fiber is 1,300 meters from White Slough today — close, but not yet connected.

Wastewater plant expansion

$60 – $120 million

The biggest cost in the package. The White Slough wastewater plant on Thornton Road is already running at its full 8.5-million-gallon-per-day capacity. The City needs to expand it anyway for projected residential and business growth, with or without a data center. Comparable Central Valley projects: Manteca spent about $130 million for a 3-MGD expansion (2020–2023). Tracy spent $80–$110 million on similar work. State Water Board permitting alone takes 18–30 months before construction can start.

Recycled water delivery

$5 – $15 million

Pipes and pumps to move recycled water from the wastewater plant to wherever the data center sits ($2–$5 million), plus deep wells to dispose of cooling-tower wastewater that the regular plant can’t handle ($5–$15 million). If the City requires the data center to use closed-loop cooling, this drops to $2–$3 million — but the developer then spends $15–$30 million more on its end.

Permits, EIR, entitlements

$3 – $6 million

An environmental impact report for the data center ($1–$3 million, with 18–30 months of preparation), changes to White Slough’s zoning if the existing utility designation needs to be broadened ($500K–$2 million), San Joaquin Valley Air Pollution Control District permits ($200K–$500K), and money set aside to defend the EIR in court if someone sues ($1–$2 million). CEQA lawsuits can add 1–3 years to a project anywhere in the schedule.

Roads and ancillary site work

$2 – $10 million

Thornton Road improvements to handle construction trucks and facility traffic ($1–$5 million). If Caltrans determines the traffic is heavy enough to require interchange work on I-5 — at SR-12 Flag City or Eight Mile Road — that runs $5–$20 million, but Caltrans pays, not Lodi. On-site work inside the data center perimeter is the developer’s cost. Lodi’s share is $2–$10 million.

Figure 1 — Site-readiness cost stack for a 10 MW White Slough data center

Source: LodiEye analysis. Component cost ranges in millions of dollars. Excludes data center building construction ($60–$150M) and the hydrogen plant funding gap. Anchored to Prime Data Centers McClellan, Manteca and Tracy WWTP expansions, and California Middle-Mile Broadband Initiative route data.

Adding It All Up: $108 to $221 Million

For a 10-megawatt facility, excluding the hydrogen plant question:

Component	Low	High	Midpoint
Electric infrastructure	$30M	$60M	$45M
Fiber and network	$10M	$20M	$15M
Wastewater expansion	$60M	$120M	$90M
Recycled water delivery	$5M	$15M	$10M
Permits and entitlements	$3M	$6M	$4.5M
Roads and ancillary	$2M	$10M	$6M
Total site readiness	$108M	$221M	$165M

The range is wide because two of the biggest pieces — the substation and the wastewater plant — depend on choices that haven’t been made yet, and on market conditions that are still moving (transformer lead times, construction costs, state permitting complexity). The middle of the range, around $150 to $170 million, is the most realistic expectation.

The cost shape changes depending on how big the data center is. Below 5 megawatts, the City is paying for utility upgrades it would probably need anyway for general business growth. Above 12 megawatts, the City is paying to fundamentally remake Lodi Electric Utility — work it wouldn’t do otherwise. Ten megawatts sits inside the second zone. Whatever Lodi spends to support a 10-megawatt facility can’t easily be repurposed for other industrial customers at a fraction of the cost.

How Long This Would Take

The earliest a 10-megawatt data center could be running at White Slough is 2031. The realistic date, accounting for normal delays, is 2033. If anything significant slips — and something usually does — 2034 or later.

The longest single item in the schedule is the transformer for the new substation. Right now, lead times worldwide are 2 to 4 years just to get the equipment delivered after it’s ordered. Larger transmission-class units run 3 to 5 years. NCPA, the Sacramento Municipal Utility District, and PG&E are all waiting in the same queue with everyone else. Any timeline that puts a Lodi data center operating before 2031 either requires the City to pre-order transformers before a tenant is signed (which means Lodi takes the risk if the deal falls apart), or requires the developer to deliver the transformer (which means losing control of the schedule).

Figure 2 — Critical path timeline for White Slough site readiness

Source: LodiEye analysis. Workstream durations and dependencies for a 10 MW project. The substation transformer lead time and wastewater plant permitting+construction are the critical-path items. A smaller 2-to-5 MW facility compresses to approximately five years; CEQA litigation can add 1–3 years anywhere in the schedule.

Workstream Sequencing

Phase	Years	Workstream Activities
Phase 0 — Feasibility	2026 Q2–Q4	Council decision on whether to pursue site readiness. Detailed feasibility study with engineering. Request for information to qualified data center operators. NCPA agreement review for transmission and generation contracting implications.
Phase 1 — Pre-Development	2027	Hydrogen plant funding outcome resolved (gating). EIR scoping for both data center and wastewater. Specific plan amendment if required. Substation engineering and PG&E interconnection application filed. Wastewater plant capacity study and permit application.
Phase 2 — Permitting and Procurement	2028	EIR certification. Substation transformer order placed (long-lead procurement). Wastewater EIR and Title 22 permits complete. Fiber design and carrier partnership negotiations. Developer RFP issued.
Phase 3 — Construction Start	2029	Wastewater construction begins. Fiber construction begins. Developer development agreement negotiation.
Phase 4 — Major Build	2030	Transformer delivery and substation construction. Wastewater expansion completion. Fiber commissioning. Development agreement execution.
Phase 5 — Tenant Build	2031	Substation commissioning. Data center construction begins. Hydrogen plant operational if proceeding on this path.
Phase 6 — Operational	2032–2033	Data center construction completes. Go-live.

All of that assumes nothing major slips. Realistically, plan on 2033 to 2034 for the 10-megawatt scenario. A smaller 2-to-5-megawatt facility — which doesn’t need the dedicated substation — could be running by 2030 or 2031. A larger 12-to-25-megawatt facility takes longer because of additional electric and wastewater work.

What Has to Happen, in What Order

Five things have to be settled in sequence. Each one has to be resolved before the next can really proceed.

1. Hydrogen plant funding outcome. Whether the data center pitch includes hydrogen power, or just natural gas and grid power. That changes the kind of operator who would be interested, the marketing story, and possibly how electricity gets contracted. As of May 2026, the Department of Energy’s $35 million cancellation is still unresolved. Representative Harder’s effort to get it restored hasn’t yielded a public answer.
2. The NCPA contract structure. Lodi Energy Center’s 296 megawatts are committed to NCPA member cities through long-term contracts. Lodi’s share is about 30 megawatts. A data center can’t just “use” Lodi Energy Center output. Either NCPA negotiates a new contract for the facility (could take years), or Lodi reallocates its existing share (other member cities have to agree), or the data center connects to PG&E’s grid instead. Each path has different costs, timelines, and politics.
3. PG&E’s queue for new grid connections. If the data center goes the PG&E route instead of NCPA, the wait time in the Central Valley for a new connection of this size is over four years right now. That’s an outside-Lodi delay no local decision can speed up.
4. State Water Board permits. The wastewater expansion, the recycled water reuse, and any deep injection wells all need state permits, each with its own timeline. End-to-end, sequential review can take 24 to 36 months.
5. CEQA review. The data center needs an environmental impact report. So does the wastewater expansion. So does the hydrogen plant, if it proceeds. Combining the reviews is more efficient but creates one big lawsuit target. Splitting them adds time but spreads the legal risk.

What Could Go Wrong

The major risks to a White Slough data center project, in approximate order of dollar impact on Lodi:

Risk	Magnitude	Likelihood	How to Mitigate
Transformer delays or price jumps	$15–$30M	High	City pre-orders the equipment, or makes the developer buy it
Wastewater plant cost overruns	$18–$36M	Moderate–High	Build in stages; charge developer based on actual wastewater volume
Environmental lawsuit delays	$0.5–$2M + 1–3 yrs	Moderate	Solid EIR documentation, early community engagement
Hydrogen plant delayed or canceled	Marketing pitch weakens	High	Structure data center deal independent of hydrogen project; treat hydrogen as upside
Data center operator backs out	Stranded investment	Moderate	Developer deposits with forfeiture; reimbursement tied to milestones
Costs pushed onto residential utility bills (Hillsboro drift)	$2–$8/mo/customer over 10 yrs	High without rules	Separate billing class for >5 MW customers, full cost-of-service rate
Other NCPA cities object	Project blocked or restructured	Low–Moderate	Early NCPA member consultation; contract review
New state rules on data centers	Compliance cost varies	Moderate	Track SB 57, SB 886, and CPUC proceedings

The single most consequential risk is cost-shifting to residential utility bills — the Hillsboro drift. It has the lowest probability if Lodi puts the right deal terms in place, but the dollar impact on Lodi families over a decade could match the entire revenue benefit of the project. That’s the risk every deal term in this article is designed to prevent.

Two Roads: Quincy and Hillsboro

Two cities started in similar places and ended up in very different ones. Quincy, Washington and Hillsboro, Oregon both began with utilities and electric capacity that looked a lot like Lodi’s today. Twenty years later, Quincy residents pay roughly one-third the property tax they did before data centers arrived. Hillsboro residents pay roughly 50 percent more for electricity than they did five years ago.

What drove the difference is specific and reproducible — and the same mechanism would apply in Lodi, even at much smaller scale.

Quincy applied three disciplines

Separate rate class. Each data center customer was placed on its own rate schedule covering full cost of service — power, distribution, substation amortization, contracts.

Up-front infrastructure payment. Substations and distribution were paid for by developers before construction, not amortized into general utility rates.

Multi-decade payment commitment. Minimum-payment contracts transferred demand risk to data centers, not residents.

Result: residents benefited financially from the data center presence.

Hillsboro applied none of these

Aggregate rate class. Data centers paid bulk rates with cost-of-service recovery aggregated across customer classes.

Rate-based infrastructure. Utility-built infrastructure was rate-based, recovered from all customers including residential.

No minimum-payment commitments. Utility carried demand risk; if data centers shrink or leave, residents pay.

Result: residential electricity rates rose roughly 50 percent over five years.

Hillsboro didn’t decide to make this happen. It happened because the rules that separate data center costs from regular customer bills were missing. Once it starts, it’s hard to reverse — by then, regular customers have been quietly paying for data center infrastructure for years.

For Lodi, the same logic applies despite the much smaller scale. Even at 10 megawatts, a data center would become Lodi Electric Utility’s biggest single customer by a wide margin. Whether that ends up helping or hurting residents depends entirely on how the costs are sorted out. The presence of Lodi Energy Center doesn’t change this. What matters is whether the data center pays the actual cost of the infrastructure built to serve it.

Figure 3 — Cumulative residential customer rate impact: Quincy vs. Hillsboro outcomes

Source: Oregon Citizens’ Utility Board residential rate data; Washington Department of Revenue Data Center Workgroup; LodiEye scenario analysis. Illustrative cumulative impact on a typical residential electricity customer over 10 years, scaled to Lodi at 10 MW.

How to Make Sure Lodi Gets the Quincy Outcome

To translate Quincy’s approach into Lodi-specific deal terms, the protective conditions from the earlier LodiEye article need specific dollar figures and contract language. The ten items below pair each protection with the dollar magnitudes from the cost stack. Each is a contract requirement, not a guideline.

1. Separate billing class for industrial customers above 5 megawatts. The data center pays the actual cost of its substation (spread over 15 years), the fiber, its share of the wastewater plant expansion, and its electricity at full cost-of-service. Estimated rate: $0.11 to $0.14 per kilowatt-hour, compared to $0.16 to $0.18 for residential customers in 2025. Revenue to the City from the rate margin: $1.2 to $2 million per year.
2. Developer pays for infrastructure up front, not over time. Substation ($30 to $60 million), fiber ($10 to $20 million), and recycled water delivery ($5 to $15 million) are paid by the developer as a development fee or up-front contribution before construction begins. Total developer payment before breaking ground: $45 to $95 million. This is the single most important rule in the entire deal. If the project gets canceled or scaled back, Lodi isn’t stuck holding the bill for infrastructure built but unused.
3. Wastewater costs split between general growth and the data center. Lodi needs to expand the WPCF anyway for residential and business growth. Of the $60 to $120 million expansion, 30 to 50 percent is attributable to the data center’s incremental discharge. The data center pays that share ($18 to $60 million) through an impact fee tied to its actual measured wastewater output. The rest is covered by general rates the City would need anyway.
4. 15-year minimum payment contract with early-exit penalty. The data center commits to paying for at least 15 years. If it closes or relocates early, it pays the present value of the remaining contract plus a 25 percent penalty. Industry standard for serious operators is 15 to 20 years. California’s pending Senate Bill 886 frames similar requirements.
5. Closed-loop cooling if evaporative cooling would tap too much recycled water. If the data center’s evaporative cooling would use more than half the recycled water available at White Slough (after accounting for existing users), it has to switch to closed-loop cooling at its own expense. This protects the recycled water for the hydrogen plant, area farmers, and other future industrial users.
6. Annual public reporting and audit rights. Detailed yearly reports on water use, power use, jobs, taxes, and wastewater discharge — broken out specifically for the Lodi facility, not buried in corporate totals. The City has the right to audit. Penalties for misreporting include permit suspension.
7. Transparent capacity accounting. The amount of electric capacity reserved for the data center is recorded publicly. Other industrial customers — Pacific Coast Producers, the wineries, food processors, future Cepheid expansion — know what’s available for them. Capacity allocation doesn’t happen behind closed doors.
8. Local hiring and spending requirements. Minimum 25 percent of construction trades and 60 percent of permanent operations roles filled by San Joaquin County residents. Minimum 20 percent of operations spending — security, maintenance, services — directed to local businesses. Missing the targets means the developer forfeits negotiated incentives.
9. No diesel backup generators. Backup power has to be hydrogen-blend, grid-tied, or battery — locked into the development agreement, not just promised in the pitch. If diesel turns out to be the only practical option, the developer pays all the air-district permitting and emissions-offset costs.
10. Fiber built for the data center also serves the community. Any new fiber installed for the data center — particularly the backup ring to Stockton — has to reserve capacity for City use: schools, public safety, underserved neighborhoods. At no extra cost to Lodi.

What Lodi Would Actually Make

Under the deal terms above, a 10-megawatt data center at White Slough would generate this much for the City each year, once operating:

Revenue Source	Annual Low	Annual High
LEU rate margin (above wholesale cost)	$1.2M	$2.0M
Property tax (Lodi share after county/district)	$0.4M	$0.7M
Possessory interest tax (if on city land)	$0.2M	$0.4M
Wastewater service fees (full cost of service)	$0.4M	$0.8M
Recycled water service fees	$0.2M	$0.4M
Total annual city revenue	$2.4M	$4.3M

Meaningful, but not transformative. It matches the earlier LodiEye estimate of $2 to $3 million a year. For context: Lodi’s 2024-2025 budget projects about a $1 million annual structural shortfall over the next five years. A 10-megawatt facility under disciplined terms would cover that shortfall and leave some room for capital needs. It wouldn’t fund major new programs the way Quincy’s data center cluster eventually did — that took 60-plus facilities at Microsoft-scale.

A note on developer reaction. These deal terms aren’t designed to make developers happy. They’re designed to protect Lodi residents. Developers who walk away from these terms will tell the Council that Lodi is being “uncompetitive” compared to Texas or Virginia. That’s true — and it’s the point. Texas and Virginia are paying for their data center attractions through subsidies that ultimately fall on residents and ratepayers. Lodi’s alternative is the model where the people of Lodi do not subsidize the developer’s profit margins. A developer who can operate under disciplined terms is a developer whose economics actually pencil out — which is the only kind of project that leads to a Quincy outcome.

Connection to Lodi’s Existing Economy

A 10-megawatt data center at White Slough would be a largely standalone industrial facility. That’s a category not unusual in itself — semiconductor fabs, regional distribution centers, and most data centers operate this way. Standalone industry isn’t automatically bad. But it works differently from the food processing, viticulture, and biotech industries that anchor Lodi’s existing economy, and the difference is worth understanding before committing City infrastructure to host one.

Limited upstream linkages. Pacific Coast Producers and Lodi’s wineries buy substantial inputs locally — grapes, fruit, packaging, labor. A data center buys essentially none of those things. Its primary inputs are electricity, water, fiber bandwidth, and equipment manufactured outside the region. The deal-term local-spending protection (Term 8 above) would secure roughly 20 percent of ongoing operations spending — security, maintenance, services — for San Joaquin County businesses. The bulk of the facility’s purchasing flows out of Lodi.

Limited downstream linkages. Food processors feed into Central Valley logistics and trucking. Wineries drive tourism, hospitality, and downtown commerce. A data center ships nothing physical. Its outputs route to customers anywhere in California, the U.S., or globally, with no Lodi-side downstream multiplier.

Limited labor crossover. The 25 to 35 permanent positions are specialized — network engineers, facility technicians, security operations — drawing from a labor pool different from food processing, viticulture, biotech, or agricultural-equipment workers. A data center adds a small, separate employment base. It doesn’t expand Lodi’s existing one.

Competition for the same finite resources. This is the operational consequence. A 10-megawatt facility would draw from the same finite pools that Lodi’s current and prospective industries also draw from:

• Electric capacity. Roughly 18 percent of Lodi Electric Utility’s 2023 retail sales. Any future Cepheid expansion, a new food processor, or new winery facilities would compete for what’s left.
• Wastewater treatment capacity. A meaningful share of the expanded plant. The rest is shared between residential growth and other industrial users.
• Recycled water. Up to half the available supply under evaporative cooling — leaving less for the hydrogen plant, agricultural users, and Lodi Energy Center cooling.
• Municipal staff capacity. Planning staff, utility engineering, and Council attention dedicated to a multi-year data center process is bandwidth unavailable for other industrial attraction or for existing City operations.
• Construction labor. The 200 to 300 construction jobs during the 24-to-30-month build would compete with all other concurrent construction in the area, including the Eastside Vision project, the wastewater expansion itself, and ongoing residential and commercial construction.

Fiber is the one possible exception. The carrier-diverse fiber ring required under Deal Term 10 — with capacity reserved for City use, schools, public safety, and underserved neighborhoods — could be genuinely additive. That’s fiber infrastructure Lodi doesn’t currently have and that other businesses could share. The benefit depends entirely on the deal-term protection being enforced. Without it, the fiber serves only the data center.

The data center is one path. It would establish a new, separate, high-revenue industrial facility on land that’s currently underutilized. It would not expand or strengthen Lodi’s existing economic base. Framing it as “industrial diversification” or “tech sector growth” overstates what it would deliver. It’s a single, isolated industrial facility in a sector with limited connection to the rest of the local economy. The Council’s decision is between this path, paths that build on what Lodi already has, or some combination of both.

What the Council Should Decide Before Committing a Dollar

Several decisions need to be settled by the Council before the City spends any predevelopment money on data center site readiness. These determine whether the project goes forward at all, and the Council should make them in public, with recorded votes:

1. Is the City prepared to commit $108 to $221 million in infrastructure investment over the next 5 to 7 years to host a 10-megawatt facility? Council position, with reasons stated for the record.
2. Does the City require Quincy-style cost separation as a precondition before any data center request-for-proposals goes out? Council position, with the specific rate structure adopted by resolution before any RFP is published.
3. Does the City wait for the hydrogen plant funding to resolve before pursuing data center site readiness? Or proceed in parallel? Council position with stated trigger conditions.
4. How does the wastewater expansion get split — general city growth versus data center share? Council position with a specific formula.
5. How is Lodi Electric Utility’s available capacity allocated between a potential data center and other industrial customers? Council position with a written allocation principle.
6. What’s the plan if the developer pulls out after the City has invested in site readiness? Council position covering write-downs, alternative tenant search, and what happens to the infrastructure that’s already been built.
7. What public engagement happens during the entitlement and development agreement negotiations? Council position with specific milestones for community input.

These decisions belong in front of an RFI, not behind it. A developer responding to an RFI that’s already committed to weak deal terms will play those terms. A developer responding to an RFI that has Quincy-style rules as a precondition will either accept or walk away — both of which serve Lodi’s interest.

What City Staff Should Confirm in the Council Report

Interim City Manager Aaron Busch said at the May 6 Council meeting that staff would bring a report back in late May or early June. The following pieces of information would make that report substantially more useful:

1. Lodi Electric Utility’s actual available capacity. Today’s peak demand, contracted capacity, and how much is uncommitted for new large customers. The numbers used in this article are estimates based on typical municipal utility patterns. Actual numbers will sharpen the scenario analysis.
2. The wastewater plant’s actual current capacity. The 8.5-MGD figure is the design capacity. The plant’s real operating capacity may be lower depending on permit limits, the condition of the collection system, and treatment technology. The 2025 compliance reports would settle this.
3. Actual recycled water availability. After Lodi Energy Center, smaller turbines, agricultural users, and the planned hydrogen plant take their shares, how much recycled water is actually available for additional industrial use? This determines whether evaporative cooling is even feasible at 10 megawatts.
4. What NCPA’s joint powers agreement actually allows. Whether non-member industrial customers can contract for on-site generation at the Lodi Energy Center, and what consent process from other member cities applies.
5. PG&E’s current queue position for new connections of this size in the Lodi-Stockton area, if grid power becomes the primary path instead of NCPA.
6. The San Joaquin Valley Air Pollution Control District’s initial position on hydrogen-blend backup generation. Whether it can substitute for diesel emergency generators under current rules, or whether full review applies.
7. Lodi Public Works’ actual wastewater expansion cost estimate. The City has presumably already scoped this for general growth. That number, adjusted for incremental data center discharge, is more specific than the Manteca-Tracy comparables used here.
8. Whether any actual developer has expressed interest. Recent discussions about White Slough have been speculative — no specific operator has been publicly named. The staff report should distinguish between general interest in the site and operators who have actually done site visits or formal inquiries.

The Bottom Line

A 10-megawatt data center at White Slough is a real possibility with a specific shape. Whether it’s a good idea for Lodi depends on the comparison.

What’s on offer. $2.4 to $4.3 million a year in City revenue once operating. Roughly 25 to 35 permanent jobs paying $80,000 to $140,000. A class of industry Lodi doesn’t currently host. Co-location advantages — Lodi Energy Center adjacent, recycled water available, land available — that genuinely set White Slough apart from competing California sites.

What it would cost. $108 to $221 million in City-side infrastructure investment over 5 to 7 years. That figure is on the order of Lodi’s annual general fund budget. It would also tie up significant municipal capacity — planning staff, utility engineering, Council attention, public process — that could be directed elsewhere. Comparable resources could fund a substantial road program, fully cover the wastewater expansion Lodi needs regardless, or seed business attraction in sectors where Lodi already has competence (food processing, biotech, viticulture). The data center is one option among several.

What could derail it. Transformer lead times are long and getting longer. The hydrogen plant funding question is unresolved. Developer willingness to accept the protective deal terms is unknown. State legislation on data centers is still moving. Any of these could turn a 2033 success into a 2034 disappointment — or worse, into stranded infrastructure with no tenant to fill it.

What could go quietly wrong even if the project happens. The Hillsboro outcome — gradual cost-shifting from a data center onto regular utility customers over years — isn’t hypothetical. It happened in a city with structural advantages comparable to Lodi’s, and it happened through the absence of clear separation between data center costs and regular customer bills. That mechanism is reproducible in any utility that doesn’t put the separation rules in place beforehand.

The honest answer to “should Lodi pursue this?” is conditional. The opportunity exists. The risks are real, and some are outside the City’s control. The protections that separate a Quincy outcome from a Hillsboro outcome are well-documented and implementable, but they have to be in place before a developer arrives — not negotiated after.

The Council’s first question isn’t whether to pursue the data center. It’s whether the City is willing — publicly, with a recorded vote — to commit to the deal-term protections that keep residents from absorbing the cost. Without that commitment, residents end up paying for data center infrastructure over decades, as Hillsboro shows. With it, Lodi can either negotiate a Quincy-style outcome or step away from a project that doesn’t fit the terms. Either path serves residents; the in-between doesn’t.

About This Report

LodiEye is the investigative research arm of Lodi411.com, a citizen-run civic data and transparency platform serving Lodi, California and San Joaquin County. LodiEye is not a traditional news outlet. It does not employ professional journalists or reporters, and the people behind it do not hold journalism degrees or have professional newsroom experience. LodiEye is best understood as civic research and analysis — not peer journalism — and is not a substitute for the local and regional news organizations that do this work professionally. For traditional reporting on Lodi, San Joaquin County, and the broader region, readers are encouraged to consult the Lodi News-Sentinel, Stocktonia, The Sacramento Bee, CalMatters, and other established news outlets staffed by credentialed journalists.

This LodiEye analysis was produced using artificial intelligence tools under the direction and review of the founder. Lodi411 uses multiple AI platforms in its research and publication workflow, including Anthropic’s Claude (primarily Opus and Sonnet models) and Perplexity AI across a variety of large language models offered by each. These tools were used in the following capacities:

Source Discovery: AI-assisted search and retrieval identified primary and secondary sources covering the California Middle-Mile Broadband Initiative route geometry through San Joaquin County, the White Slough Water Pollution Control Facility, the Lodi Energy Center and the broader NCPA generation framework, the proposed Lodi Energy Center Hydrogen Project and the October 2025 federal funding cancellation, the City of Lodi’s General Plan and 2024-2025 budget, comparable Central Valley wastewater treatment plant expansions (Manteca, Tracy), comparable California substation and data center infrastructure projects (Prime Data Centers McClellan, Quest Roseville, NTT Sacramento, Nautilus Stockton), comparable city outcomes (Quincy, Prineville, Hillsboro), and California’s evolving regulatory environment for large electricity loads. Perplexity AI was used for initial source discovery and real-time data retrieval; Claude was used for deeper analysis and synthesis.

Credibility Validation: Sources were tiered by credibility, prioritizing government and regulatory documents (California Department of Technology, California Energy Commission, Northern California Power Agency, City of Lodi, U.S. Energy Information Administration, U.S. Department of Energy, California Public Utilities Commission, State Water Resources Control Board) over secondary news and industry research. The California Middle-Mile Broadband Initiative geospatial dataset was processed directly to derive route proximity values; numerical claims about distances and infrastructure positions are calculated rather than estimated. Cost figures were anchored to comparable California projects and cross-referenced where possible.

Analysis and Synthesis: Claude performed the spatial analysis against the California Middle-Mile Broadband Initiative GeoJSON dataset to derive fiber-route proximity values for the White Slough complex, the component-by-component site-readiness cost stack calculation, the critical-path timeline analysis with dependency mapping, the comparative cost-allocation modeling between Quincy-style and Hillsboro-style deal-term structures, and the revenue-to-city projections under disciplined rate-class assumptions.

Presentation: Claude assisted in drafting prose, structuring the article narrative around cost-and-dependency analysis rather than market-aspirational framing, designing the Kendo chart specifications for the cost stack, critical path timeline, and outcome divergence visualizations, and organizing the content flow from scoping question through analysis to deal-term architecture and Council decision items.

Final Review: Multiple AI models reviewed the completed draft for factual consistency, numerical accuracy, source attribution, logical coherence, and balanced presentation. Multi-tool cross-checking across Claude and Perplexity served as the principal error-reduction mechanism. Editorial judgments, attribution decisions, scenario calibration, and final publication decisions were made by the founder.

Lodi411/LodiEye believes transparency about AI use serves both readers and the broader information ecosystem. Readers who spot errors are encouraged to write editor@lodi411.com so corrections can be made.

References

1. California Department of Technology. Statewide Middle-Mile Network Map and project status data; San Joaquin County route segments (OBJECTID 74339, 74388, 74428, 74477). broadbandforall.cdt.ca.gov
2. California Department of Technology. Middle-Mile Broadband Initiative Network Development page; delivery method breakdown by county. Updated February 23, 2026.
3. California State Geoportal. MMBI Statewide Network GeoJSON dataset (bb8f61cc57fe4582b7236235e52ac7e7).
4. City of Lodi. Lodi News-Sentinel coverage of May 6, 2026 City Council shirtsleeve session on data center discussion; remarks by Council members Bregman, Craig-Hensley, Yepez, Nakanishi, and Hothi; remarks by Lodi Electric Utility Director Jeff Berkheimer and Interim City Manager Aaron Busch.
5. City of Lodi. White Slough Water Pollution Control Facility documentation; current NPDES discharge volume; design capacity 8.5 million gallons per day.
6. Northern California Power Agency. Lodi Energy Center documentation; member city share allocation; combined cycle natural gas generation capacity 296 MW with 45 percent hydrogen blend capability.
7. U.S. Department of Energy. ARCHES California Hydrogen Hub selection (October 2023) and subsequent $35 million Lodi Energy Center Hydrogen Project funding cancellation (October 6, 2025). Office of Clean Energy Demonstrations.
8. U.S. Representative Josh Harder. October 6, 2025 statement and subsequent correspondence with Energy Secretary Chris Wright regarding restoration of Lodi hydrogen funding.
9. GHD. Northern California Hydrogen Production Facility at Lodi development plan documentation. 60 MW reference plant design; 24 metric tons hydrogen per day production capacity.
10. Prime Data Centers. Sacramento campus at McClellan Park documentation; 50 MVA substation expandable to 150 MVA; carrier-neutral facility with multiple Tier 1 long haul fiber paths; 3 millisecond latency to San Jose and San Francisco.
11. Nautilus Data Technologies. Stockton 1 facility documentation; 7 MW capacity; 18-mile, 288-strand redundant fiber ring built by Utility Telecom to local carrier hotel; carrier-neutral colocation.
12. City of Manteca. Wastewater Treatment Plant expansion project documentation 2020-2023; approximately $130 million for 3 MGD capacity addition.
13. City of Tracy. Wastewater Treatment Plant expansion documentation.
14. Cushman & Wakefield. U.S. Data Center Development Cost Guide, 2025 and 2026 editions.
15. CBRE. North American Data Center Trends Report, 2025 mid-year. Identification of 5-to-10 megawatt regional segment as principal growth area outside hyperscale.
16. Washington Department of Revenue. Data Center Workgroup Preliminary Report. December 1, 2025.
17. Oregon Citizens’ Utility Board. Residential electricity rate data for Portland General Electric service territory; 2020-2025 trend.
18. City of Quincy, Washington. Property tax rate history and data center customer composition. Quincy Valley Post-Register, “Quincy Data Centers Drive Growth, Lower Levy Rates” (January 2026).
19. Crook County, Oregon. Fiscal documentation on Meta and Apple data center campuses in Prineville; fee-on-electricity-use model.
20. State of Oregon. House Bill 3546 (POWER Act, 2025); data center cost-of-service requirements.
21. California Senate Bill 886 (pending). Minimum 15-year payment commitment framework for large electricity loads.
22. California Public Utilities Commission. Senate Bill 57 (Padilla, 2025) docket; data center cost-shifting study.
23. Public Advocates Office, California Public Utilities Commission. “How Will Data Center Growth Impact California Ratepayers?” October 2025.
24. Little Hoover Commission. “Data Centers and California Electricity Policy.” March 2026.
25. San Joaquin Valley Air Pollution Control District. Permitting guidance for combustion-based backup generation.
26. California State Water Resources Control Board. NPDES permit framework for industrial wastewater discharge; Title 22 recycled water requirements; deep injection well permit framework.