That is a tight schedule for utility-scale solar at this size and leaves limited margin if interconnection review, transformer procurement, or weather events create delays

Decision Lens

Google’s energy strategy now runs two tracks simultaneously: aggressive developer acquisition and conventional long-term PPAs. The $4.75 billion Intersect acquisition in late 2025 signaled a move toward owning generation capacity; the Lupinus agreements — two PPAs covering approximately 400 MWac of Texas solar — confirm that traditional offtake remains a primary tool alongside ownership. The question worth sitting with is whether this reflects owned generation being unable to scale fast enough geographically, or whether PPAs are bridging supply gaps while the acquired pipeline matures. For energy heads with Texas load commitments, the late Q4 2027 commercial operation date is already a hard dependency on your planning calendar.

90-Second Brief

Today, google has signed two long-term PPAs with Sunraycer Renewables for the Lupinus and Lupinus 2 solar projects in Franklin County, Texas, totaling approximately 400 MWac. Construction began in late March 2026, with commercial operation targeted for late Q4 2027. The deal follows Google’s $4.75 billion acquisition of renewable developer Intersect in late 2025. Taken together, both moves reflect sustained hyperscaler pressure to secure firm generation capacity ahead of AI-driven load growth, using every available procurement instrument in parallel.

What’s Actually Happening

The Lupinus portfolio is a utility-scale photovoltaic development within ERCOT territory, a market where data center load growth has intensified competition for cleared generation positions. Google’s director of energy and power, Will Conkling, framed the deal explicitly around grid reliability and affordability in Texas — language that points to a specific regional load obligation rather than a portfolio-wide REC volume play.

Construction commencement in late March 2026 with a late Q4 2027 target allows roughly 20 months to first power. That is a tight schedule for utility-scale solar at this size and leaves limited margin if interconnection review, transformer procurement, or weather events create delays.

What the source confirms: two signed PPAs, a named developer, a defined project site, and a targeted commercial operation date. What it does not confirm: pricing, contract tenor, curtailment provisions, or the interconnection queue status of the projects.

Why It Matters for Global Heads of Data Center Energy?

Texas is the most contested data center energy market in North America. ERCOT’s energy-only market structure means every PPA carries direct basis risk: a solar offtake without firming capacity leaves operators exposed during high-demand, low-generation periods — exactly the conditions that coincide with peak data center load. Google’s decision to sign PPAs through an emerging developer rather than wait for Intersect output suggests that owned generation cannot yet deliver the geographic reach or timeline speed required to cover active load commitments.

For energy heads benchmarking their own procurement stack, the critical question is whether your Texas PPA book is sized by site-specific load — or whether you are relying on regional portfolio averaging to meet Scope 2 or 24/7 CFE targets. The Lupinus transaction also shifts the competitive landscape for developer access: if hyperscalers are contracting 400-MW blocks with mid-tier developers, shovel-ready ERCOT projects across the developer spectrum are being absorbed faster. Waiting for tier-one IPP availability in Texas carries an increasingly real opportunity cost.

The Forward View

Google has now layered two procurement instruments — owned development via Intersect and contracted generation via long-term PPA — and the operational question is how these interact as both mature simultaneously. Once the Intersect pipeline produces projects at scale and with geographic coverage, incremental PPAs may shift from necessity to optionality. Until then, the 2027 commercial operation window is a firm dependency, not a hedge.

For the broader ERCOT market, expect continued hyperscaler PPA activity through 2026 as operators try to synchronize generation availability with data center commissioning timelines. Developers holding late-2027 or 2028 commercial operation dates with cleared interconnection positions in ERCOT are likely seeing accelerated inbound interest from offtake buyers. The risk of over-contracting — if AI workload projections moderate — is real, but there is no public evidence that procurement behavior is currently pricing in that downside scenario.

What We’re Uncertain About?

  • PPA financial terms and curtailment structure: No pricing, tenor, or curtailment provisions have been disclosed. Without these, it is not possible to evaluate whether the Lupinus agreements are cost-competitive relative to ERCOT spot markets or alternative offtake structures available to peers today.

  • Intersect pipeline and Texas coverage: Google paid $4.75 billion for Intersect, but the geographic distribution of Intersect’s project pipeline — and whether it includes near-term Texas capacity — has not been publicly confirmed. Whether Lupinus was contracted precisely because Intersect lacks imminent Texas output remains an open question that would materially change how peers interpret this move.

  • Project execution risk on the 2027 timeline: The source confirms construction has started but does not confirm interconnection queue position or transformer procurement status. At 400 MWac, transformer lead time alone represents a schedule risk that could compress the margin to the late Q4 2027 target. What would resolve this: a public interconnection status filing or equipment procurement confirmation from Sunraycer.

  • Developer counterparty depth: Sunraycer is not a volume-established tier-one IPP. On a project of this size, developer execution capability — EPC contracting, financing close, and O&M readiness — matters to delivery certainty. The source does not provide sufficient basis to assess counterparty risk, and peers evaluating similar developer relationships should treat this as an open variable.

One Question to Bring to Your Team

If a hyperscaler with a $4.75 billion owned-developer asset is still signing long-term PPAs through a mid-tier developer to cover a specific region, what does your current procurement framework say about when to contract externally versus waiting for owned or partner generation output — and do your Texas load commitments have a confirmed generation backstop dated before your next commissioning window?

Sources

  • Solarpowerworldonline — Google signs PPAs on 400-MW Texas solar project portfolio (Link)

Tags