The Cooling Crisis: Manufacturing Boom Amid Demand Uncertainty

3. Supply Chain Dependencies and Cascading Effects

3.1 Thermal Management Industries Beyond Data Centers

The rush to build cooling capacity for data centers is expanding into equipment categories and supply chains that serve critical non-datacenter industries. Major industrial and pharmaceutical operations depend on the same chiller and thermal management equipment being built for AI infrastructure:

3.2 Competing Supply Chains for Shared Components

The cooling equipment expansion creates potential supply chain conflicts at the component level. Critical bottleneck elements include:

• Refrigeration Compressors: The global refrigeration compressor market was valued at $19.2 billion in 2024 and is growing at 4.7% CAGR to reach $26.7 billion by 2030. These are used across pharmaceutical, food processing, data center, and HVAC applications simultaneously.
• Heat Exchangers and Core Components: Semiconductor chips, compressors, and heat exchangers remain in high demand with long lead times. Manufacturers report needing to plan months in advance for critical materials.
• Refrigerant Supplies: Regulatory transitions to low-GWP (Global Warming Potential) refrigerants are disrupting supply chains. As of January 1, 2026, installation of HVAC systems with GWP greater than 700 becomes illegal, forcing industry-wide retooling. Current shortages of R-454B refrigerant cylinders are already contributing to residential HVAC sales slowdowns.
• Raw Materials: Copper, aluminum, and steel price volatility affects all manufacturers simultaneously. Inflation and geopolitical factors have increased costs, forcing long-term contract lockups that limit flexibility.

3.3 Current Supply Chain Stress Indicators

The HVAC and refrigeration industry is already reporting significant constraints:

• Equipment shortages and longer lead times across semiconductor chips, compressors, and heat exchangers
• Global shipping constraints due to port congestion, labor shortages, and increased freight costs
• Volatile raw material pricing creating unpredictable cost structures
• Manufacturing retooling for refrigerant standards occurring simultaneously with datacenter capacity buildout

4. Demand Uncertainties and Slowdown Risks

4.1 Growth Deceleration Indicators

Despite the current expansion mania, several indicators suggest data center demand growth may not sustain at projected rates:

Key risk factors include:

• Overcapacity Recognition: Ares Management (a major data center investor) explicitly warned that "historically when this much capacity comes online, some of it has to be marginal," noting that "these trends tend to lead to overbuilds in certain places."
• AI Monetization Uncertainty: The underlying assumption driving data center expansion is rapid AI revenue growth and adoption, but commercial ROI from large-scale AI deployments remains unproven for many applications.
• Competitive Supply: As hyperscalers (Microsoft, Google, Amazon) develop captive data center capacity and alternative asset managers (Blackstone, Brookfield, Apollo, Ares) build competing infrastructure, pricing power for data center services may erode, reducing customer demand for incremental capacity.
• Energy and Grid Constraints: Data centers face increasing power availability limitations. Microsoft's internal forecasts show Azure capacity constraints persisting through mid-2026 in key markets, indicating that supply constraints (not demand) may become binding.
• Labor Market Pressures: Construction unemployment has reached 3.2%, tied for a record low, indicating tight labor availability and rising wage pressures that could slow construction completion timelines and increase costs.

4.2 Market Saturation Signals

Emerging research suggests the market may be entering saturation phase earlier than currently assumed:

• Industry analysts already discussing "overcapacity and market saturation" in data center reports
• Fortune reporting on "risks of overbuilding" as capital floods into AI infrastructure
• McKinsey analyzing stranded asset risks if data center capacity does not materialize as projected
• World Economic Forum highlighting $3.3 trillion in potential cumulative annual losses if data centers encounter heat and climate challenges

5. Scenario Analysis: Demand Slowdown Impact

5.1 Moderate Slowdown Scenario (2026-2027)

Assumption: Data center construction moderates to 5-10% annual growth instead of 28.3% current CAGR, driven by AI monetization disappointments and overbuilding corrections.

Impact on Cooling Manufacturers:

• Excess production capacity across XNRGY, AAON, Modine, and competitors idled or sold at discounted rates
• Previously secured orders (like AAON's $174.5 million contract) become the primary revenue source
• Manufacturing facility utilization falls below breakeven for newly constructed plants
• Workforce reductions and facility consolidations become necessary
• Capital write-downs as planned expansions become uneconomical to complete

Impact on Dependent Industries:

• Positive: Component availability improves for pharmaceutical, food processing, and chemical manufacturing as datacenter demand pressure eases
• Negative: Thermal management equipment pricing rises as manufacturers attempt to recover fixed costs across smaller customer bases
• Mixed: Smaller suppliers and regional manufacturers serving non-datacenter industries may face raw material cost increases if global suppliers consolidate capacity

5.2 Severe Contraction Scenario (2027-2029)

Assumption: Major AI projects underperform expectations, leading to 30-50% reduction in data center expansion plans and capacity cancellations.

Financial Impact on Manufacturers:

• Substantial asset impairment across newly built facilities
• Stranded capacity becomes "the elephant in the data center" multiplied across supply chain industries
• Bankruptcy risk for smaller thermal management suppliers without diversified customer bases
• Market consolidation as well-capitalized firms acquire distressed competitors
• Estimated losses: $10-15 billion+ across cooling equipment, HVAC, and related thermal management industries

Impact on Dependent Industries (Severe Scenario):

• Pharmaceutical manufacturers face reduced equipment options as suppliers exit markets or consolidate
• Long-term equipment supply agreements become critical as manufacturing capacity contracts
• Regional manufacturers face supply chain fragmentation and reduced economies of scale
• Food and beverage cold chain could experience temporary supply disruptions if companies exit regional markets

5.3 Fragmentation Risk: The Supply Chain Paradox

A particularly insidious scenario involves moderate demand slowdown combined with manufacturing consolidation:

6. The Stranded Capacity Problem

6.1 Quantifying Excess Manufacturing Infrastructure

Current expansion plans total approximately:

• XNRGY: Nearly 1 million square feet of manufacturing footprint
• AAON/BASX: 787,000 square feet in Memphis plus Longview facility expansion
• Modine: Significant expansion at Mississippi facility (completion target late 2026)
• Henkel: 70,000 square foot facility in South Dakota
• Evapco, Colmac, and other competitors: Expansion details less publicized but occurring

This represents millions of square feet of manufacturing capacity specifically built or retrofitted for data center cooling equipment production. At 3,000 chillers annually from XNRGY's Mesa 2 facility alone, with similar capacity targets at competing manufacturers, the market is building for 15,000-20,000+ chiller units annually across manufacturers, plus massive quantities of ancillary cooling systems.

6.2 Fixed Cost Burden

The capital investments in cooling equipment manufacturing carry substantial fixed cost burdens:

• Modine's $38 million investment must be amortized regardless of utilization rates
• XNRGY's $10 million environmental testing chamber is dedicated infrastructure
• Worker training programs, equipment installation, and facility preparation represent sunk costs
• Long-term supplier contracts for components and raw materials lock manufacturers into commodity costs

If utilization falls from planned 80-90% to 40-50%, manufacturers face acute profitability pressures. Unlike many industries, manufacturing facilities cannot be quickly repurposed—a 330,000-square-foot specialized chiller production facility has limited alternative uses.

7. Broader Industrial and Economic Implications

7.1 Supply Chain Concentration Risk

The cooling equipment expansion creates systemic vulnerability in industrial supply chains:

• Pharmaceutical Industry Risk: Temperature-critical drug manufacturing depends on specialized chillers. Supply chain fragmentation following datacenter-driven consolidation could disrupt drug production and affect pharmaceutical supply chains serving critical medicines.
• Food Security Risk: Cold chain infrastructure depends on reliable compressor and chiller availability. Fragmentation or price spikes could increase food spoilage, reduce supply chain efficiency, and increase food prices.
• Chemical Production Risk: Process cooling for chemical synthesis is non-negotiable. Supply constraints or price increases directly pass through to chemical product costs.
• EV Production Risk: As thermal management systems become critical for battery and vehicle cooling, any supply chain disruption affects EV production scaling.

7.2 Employment and Regional Economic Exposure

The manufacturing expansion creates concentrated employment in specific regions:

• Mesa, Arizona: 500+ direct jobs from XNRGY
• Grenada County, Mississippi: 450+ jobs from Modine
• Longview, Texas: Significant AAON expansion (exact employment not disclosed)
• Memphis, Tennessee: Planned AAON facility (employment TBD)
• Brandon, South Dakota: Henkel facility expansion

These are high-value manufacturing jobs, but they are concentrated in facilities and companies entirely dependent on sustained data center growth. Regional economies that have built expectations around these employment opportunities face significant risk if demand contracts.

7.3 Capital Allocation Inefficiency

The cooling equipment expansion exemplifies a broader problem in capital markets: when asset classes become fashionable, capital floods in, creating overcapacity that inevitably leads to stranded assets and below-market returns.

8. Market Dynamics and Competitive Pressures

8.1 The Capital Overdose Phenomenon

Alternative asset managers (BlackRock, Temasek, Brookfield, Apollo, Ares, and others) are committing massive capital to data center development. This creates powerful incentives for supply chain vendors to expand capacity regardless of fundamental demand assessment:

• Larger data center investors create steady customer demand for cooling vendors
• Cooling vendors have limited leverage to resist investing in additional capacity when major customers appear committed to expansion
• Financing for manufacturing expansion becomes available as financial institutions perceive low-risk infrastructure assets
• First-mover advantage in production capacity creates competitive pressure for followers to expand as well

This dynamic resembles classical overinvestment cycles: each participant rationally responds to current demand signals and competitive pressure, but collectively they create excess supply that no individual participant can prevent.

8.2 AAON's Strategic Position

AAON's approach illustrates both opportunity and risk:

• $174.5 million in orders from a single customer provides revenue certainty for capital investments
• The company emphasizes having "considerable excess capacity," suggesting they are confident in expanded utilization
• Multiple facility expansions (Longview existing facility, new Texas facility, Memphis 787,000 sq ft acquisition) represent aggressive growth betting
• CEO statement about "fraction of a much larger pipeline" suggests confidence in multi-year demand visibility

However, this strategy also concentrates risk: if the "pipeline" doesn't materialize or customer commitments weaken, AAON faces rapid capacity obsolescence.

8.3 The Financing Paradox

Significant institutional capital is backing cooling equipment expansions specifically because data center infrastructure is perceived as "lower risk" than direct AI equity investments. This creates a dual problem:

• Capital is flowing into cooling equipment manufacturing based on data center demand assumptions
• These same assumptions may be wrong if AI ROI disappoints
• Yet cooling equipment is funded as "boring infrastructure" with stable returns expectations
• When data center demand moderates, the "low-risk" thesis evaporates rapidly

9. Regulatory and Policy Considerations

9.1 Refrigerant Transition Risks

The mandated transition to low-GWP refrigerants creates a scheduled supply chain disruption:

• January 1, 2026 Ban: Installation of HVAC systems with GWP > 700 becomes illegal in the U.S.
• Manufacturing Complexity: Factories retooled assembly lines months ago to meet January 1, 2025 manufacturing ban for R-410A
• Supply Chain Impact: Compressor supply chains adjusted for new refrigerant compatibility, but transition risks include:
• Supply shortages if manufacturers overestimate demand for new refrigerant-compatible equipment
• Higher costs for compliant equipment during transition period
• Performance differences requiring system redesigns
• Concurrent Risk: This mandated transition occurs simultaneously with maximum datacenter capacity expansion demand, creating potential supply conflicts

9.2 Power Grid and Energy Policy

Data center cooling expansion is intertwined with grid infrastructure and energy policy:

• Data center electricity demand projected to double within five years according to JLL and others
• European data center power demand forecast to grow from 96 TWh in 2024 to 236 TWh by 2035
• Grid capacity constraints in specific regions (Virginia, Texas) are already binding, causing Microsoft to restrict new subscriptions
• Renewable energy transition creates competing demand for transmission capacity
• Water availability constraints may limit cooling options in water-scarce regions

Cooling equipment manufacturers are betting on energy infrastructure that may not materialize as assumed, particularly if grid modernization underperforms projections.

10. Key Risks and Conclusions

10.1 Synthesis of Systemic Risks

Risk Ranking (by potential impact):

10.2 Key Takeaways

For Investors: Cooling equipment manufacturers are expanding based on demand signals that may not prove durable. Equipment capacity built for 2026-2028 could face 30-50% utilization rates if data center demand moderates. Risk/return calculations should incorporate severe downside scenarios.

For Supply Chain Participants: The pharmaceutical, food processing, and chemical manufacturing industries should assess their thermal management equipment supplier concentration risk. Supply chain fragmentation following potential datacenter-driven consolidation could increase costs and reduce reliability.

For Policymakers: The convergence of refrigerant transition requirements and maximum datacenter expansion demand creates a scheduled supply chain stress test. Advance planning for potential supply conflicts in compressor, heat exchanger, and chiller availability would be prudent.

For Regional Economies: Areas hosting new cooling equipment manufacturing (Mississippi, Arizona, Texas, Tennessee, South Dakota) have built employment expectations around sustained data center growth. Contingency planning for potential demand moderation would reduce regional economic vulnerability.

10.3 Conclusion

The cooling infrastructure expansion driven by AI data center buildout represents one of the most synchronized industrial investment cycles in recent memory. Billions of dollars of manufacturing capacity are being constructed on the assumption of sustained 25%+ data center growth extending through 2030. However, the underlying demand drivers—AI monetization and adoption—remain largely unproven at the scale required to justify this capacity buildout.

The risks extend well beyond data center operators. Pharmaceutical manufacturers, food processors, and chemical producers depend on the same supply chains being mobilized for data center cooling. If demand moderates, the resulting excess capacity, supply chain fragmentation, and pricing pressures will cascade through multiple industries critical to public health, food security, and global commerce.

This analysis does not argue that data center growth will necessarily disappoint. Rather, it highlights that current investment decisions in cooling equipment manufacturing embed significant demand risk without proportional hedging mechanisms. History suggests that when industrial capacity expansion becomes synchronized and fashionable, overcapacity and stranded assets inevitably follow.