The reason most Mexico advisors do not publish what follows is straightforward: the news is bad, and the people paying for the advisory want to be told the news is good. We disagree with that approach. Executives discover these facts after they commit, and that is the moment the relationship goes bad. We would rather lose a prospect at the read than lose a client at the audit. So this article is the read.
Two structural constraints are reshaping site selection in Chihuahua right now. Neither is a five-year forecast. Both are documented in the news cycle of the last 60 days. One is water. The other is grid capacity. They affect different operations differently. Both have to be in the model.
Water — what is actually happening
As of March 2026, 42 of Chihuahua’s aquifers are in deficit due to severe overexploitation [1]. That is the state’s own data, published in El Heraldo de Chihuahua. The figure is not a projection. It is a present-day audit of the basins that supply the state’s industrial corridors, agricultural belt, and municipal systems. The threshold for “deficit” is the point at which extraction exceeds recharge — once a basin crosses it, the only path back is reduced extraction, not better rainfall.
In parallel, the state has formally acknowledged that it will not meet its 2026 obligations under the 1944 US–Mexico water treaty [2]. The Conchos River — the principal Mexican contributor to the treaty’s Rio Grande commitments — delivered 0.838 Mm³ in the 2024–2025 cycle, far below historical norms. The downstream effects are political (the treaty is now in active renegotiation [3], [4]) and physical (reservoirs and rivers are depleted enough that the state has nothing to give).
The third leg is rainfall. Most of 2024 and 2025 produced no meaningful precipitation across the state. Climate-zone analysis [6] places the entire US–Mexico border region in a drought regime that is structural, not cyclical.
“Site selection in Chihuahua now has to model 5- and 10-year aquifer projections, not just current concession availability.”
Water — the operating implication
Three categories of operation should think about Chihuahua water differently.
Water-light operations — assembly, contract electronics, automotive harness manufacturing, light medical-device assembly — carry minimal water risk. Most of the day-to-day water draw is sanitary and process-cooling adjacent. Chihuahua remains a strong fit. The cluster depth, US proximity, and labor profile dominate the decision.
Water-medium operations — automotive paint shops with closed-loop reuse, electronics fabrication wash lines, mid-volume food processing — should expect a documented water plan as a precondition to siting. That plan needs three things: (1) park-level concession status under current SAT and CONAGUA regulations, (2) a quantified industrial reuse strategy, and (3) a contingency for the regulatory shift currently being debated in federal congress that would prioritize domestic over commercial use in shortage zones.
Water-heavy operations — semiconductor fabs, high-volume beverage and food processing, primary metals, pulp and paper — should not site in Chihuahua in 2026 without a serious infrastructure investment built into the project economics. Either the project pays for the water (private treatment, deep wells, alternative basins, transferred concessions) or it picks a different region. Bajío and the Yucatán peninsula have moderate-risk profiles; the central altiplano is variable; the northeast (Monterrey) has improved with the El Cuchillo II aqueduct but remains classified high-risk.
Energy — the smoking gun
This one is short and clear.
In March 2026 El Diario de Chihuahua reported that CFE’s inability to supply grid capacity is actively halting new industrial developments in the state [5]. This is not a forecast. It is not a theoretical risk. It is a current event. Projects that have signed leases are sitting unenergized, waiting on substation capacity that has not yet been built. Industrial parks in Cd. Juárez and the Manuel Ojinaga corridor are quoting connection timelines measured in years, not months.
The structural cause is national, not regional. Mexico’s transmission network was designed for 2018 demand levels. Demand has grown 3.4–3.5% per year. The transmission network has expanded 0.09–0.10% per year. Compounding gap. The reserve margin has compressed every year since.
“Plan for the published CFE timeline as a floor, not a ceiling.”
Energy — the structural cause
The federal response is real but multi-year. The March 2025 Electricity Sector Law restructured the institutional architecture; the announced pipeline includes 12 new power plants for 5,840 MW across nine states; and the “poles of development” concept concentrates investment in designated industrial geographies. Solar systems under 0.7 MW were classified as exempt generation, meaning a self-generation rooftop layer can be installed without a CRE permit.
None of these moves cure the 2025–2027 window. They cure the 2028–2030 window. For an operation breaking ground in Chihuahua during the 2026 build cycle, the operating assumption has to be that the CFE backbone is constrained and the project will need a self-supply layer to insulate against connection delay.
What this means for an operation breaking ground in Chihuahua
The take-home is not “do not site in Chihuahua.” The take-home is “model the constraints before signing the lease.” Three rules of thumb that we apply on every engagement:
Pick the park on its existing energization, not the brochure. Operating parks with documented capacity to the substation level are a different asset class than greenfield parks waiting on CFE. Pay the rent premium. The park-selection error is the most expensive mistake we see new entrants make.
Bias water-light. If the operation has any choice in process design — water-cooled vs. air-cooled, batch vs. continuous wash — bias every decision toward lower water intensity. Both for the project economics today and for the regulatory exposure tomorrow.
Build the contingency into the schedule. Add 6–12 months to whatever the park developer or CFE quote says. That is not pessimism; it is the empirical observation across our last 18 months of project starts.
The closing
Companies that absorb this read into their project economics succeed in Chihuahua. The state’s cluster depth, labor continuity, and US-border proximity remain real and valuable. The water and energy story is a constraint, not a verdict. The companies that ignore it find out about it 14 months in, when the substation is not built and the well runs short and the project is already too sunk to relocate.
If you are evaluating Chihuahua in 2026, the question to bring to a discovery call is not “is the state still a good idea?” — that depends on the operation. The question is: “given this operation’s water and power profile, what is the right park, the right service structure, and the right contingency budget?” That conversation we are happy to have.