AI data centers require massive amounts of electricity, but securing that power is no longer a simple utility transaction.
Developers may need to evaluate a mix of:
Solar Harmonics helps California data center developers look at these options as part of one coordinated power strategy.
The goal is not simply to buy electricity.
It is to understand:
Where will the power come from?
How dependable is that supply?
What will it cost over time?
How quickly can it become available?
How does it support renewable energy goals?
What additional energy resources may be available around the project?
A strong procurement strategy should answer all of those questions before the project becomes dependent on a single path to power.
For traditional commercial development, power procurement may be addressed after a site has already been selected.
For an AI data center, that approach can create major risk.
Electricity availability can influence:
A site can have excellent land, fiber, zoning, and transportation access and still become difficult to develop if sufficient electricity cannot be secured on the required timeline.
That is why power procurement should be treated as an early development decision rather than a late-stage operating decision.
The lowest-cost electricity option is not always the most valuable.
A power source may appear attractive on price but still create challenges if:
For AI data center developers, the better question is:
What combination of power sources provides the strongest balance of cost, availability, reliability, timing and long-term flexibility?
That is the foundation of a more resilient procurement strategy.
Large AI data centers may rely on several different sources of electricity over the life of the project.
The appropriate mix depends on the location, serving power provider, utility capacity, project timeline, and long-term energy objectives.
For many projects, the local utility will remain the primary electricity provider.
Utility service can offer established infrastructure and reliable access to the wider grid, but the available capacity and timeline for major new loads can vary significantly by location.
Important questions may include:
Utility power should be evaluated not only as a source of electricity, but as a development constraint and long-term business relationship.
Some California cities operate or participate in local power systems.
Municipal power can create a different procurement environment because electricity supply, infrastructure planning,g and economic development may be more closely connected.
Depending on the market, a municipal power provider may offer different:
This can create opportunities that should be understood early in site evaluation.
A power purchase agreement, or PPA, can allow a data center or associated entity to contract for electricity from a specific generation resource over a defined period.
PPAs can support long-term energy planning and renewable energy objectives.
They may involve:
The structure of a PPA can vary significantly, so the contract needs to be evaluated in the context of the project’s actual load, utility environment,nt and risk profile.
Some projects may develop energy resources directly on the data center campus.
Potential options may include:
Onsite resources can provide additional control and flexibility, but available land and system scale can limit how much of the data center’s total demand can be supplied from the property.
The procurement strategy can also include energy resources located outside the data center campus.
That may involve:
These resources can complement the primary supply strategy and expand the amount of renewable generation available to the project.
For many AI data centers, the strongest strategy may ultimately look less like one electricity contract and more like a portfolio:
Utility or Municipal Power
+
Long-Term Procurement
+
Renewable Generation
+
Solar + Battery Storage
+
Distributed Energy Resources
+
Future Expansion Capacity
The exact mix will differ by project.
The important point is to avoid treating power procurement as a single transaction when the energy requirements are this large and this consequential.
Solar Harmonics does not begin by recommending a particular electricity contract, PPA, solar project, or battery system.
We begin by understanding the project.
The objective is to determine what the data center actually needs before evaluating how those needs should be supplied.
The first step is establishing the project's energy profile.
That may include:
- Proposed location
- Initial electricity load
- Future electricity load
- Development phases
- Desired energization date
- Reliability requirements
- Sustainability commitments
- Expansion plans
- Existing power agreements
This establishes the demand side of the strategy.
Next, we determine how electricity works in the local market.
That may involve:
- Investor-owned utility
- Municipal power provider
- Community Choice Aggregator
- Existing generation resources
- Local energy programs
- Interconnection environment
- Known infrastructure constraints
The purpose is to understand who controls the various parts of the power path.
The project may have several potential sources of electricity.
These could include:
- Utility service
-Municipal power
- Long-term supply contracts
PPAs
- Renewable power
- Onsite generation
- Offsite generation
Each option is evaluated in relation to:
- Availability
- Timing
- Reliability
- Cost
- Contract structure
- Expansion capacity
- Sustainability goals
The primary power supply does not have to be the entire energy strategy.
Solar Harmonics also evaluates where additional resources may create value.
Potential resources can include:
- Onsite solar
- Offsite solar
- Battery storage
- Distributed commercial solar
- Local generation
-Municipal energy projects
These resources may complement the primary supply rather than replace it.
This is where the Solar Harmonics approach expands beyond conventional procurement.
We look at the energy opportunities surrounding the project.
That may include:
- Commercial properties
- Industrial facilities
- Warehouses
- Municipal buildings
- Parking areas
- Available land
- Potential battery sites
The objective is to determine whether additional local energy resources could become part of the broader procurement strategy.
The procurement structure also has to work over time.
That means asking:
- Can capacity expand?
- Can new generation be added?
- Can additional storage be integrated?
- Does the contract accommodate changing loads?
- Are there alternative resources if conditions change?
- Can the strategy evolve with the project?
Step 7: Build the Power Portfolio
The final result may not be one recommendation.
It may be a coordinated portfolio that includes:
- Primary Power Supply
- Long-Term Procurement
- Renewable Energy
- Solar Generation
- Battery Storage
- Distributed Energy Resources
- Future Expansion Capacity
The exact combination depends on the project.
For AI data center developers, one of the most important procurement questions is no longer simply:
“What is the cost of electricity?”
It is:
“When can the electricity actually become available?”
A project may have access to an attractive long-term power source, but if the infrastructure required to deliver that electricity takes years to build, the development timeline may be affected.
That is why time to power has become a core part of data center energy strategy.
The first question is how much power is actually available at the site today.
That includes:
A site that appears strong on paper may still require major infrastructure investment before it can support the project’s full load.
Large new loads may require:
Each of these can affect the timeline.
The procurement strategy should account for both the electricity contract and the infrastructure required to make that electricity usable.
Some projects may not need their full ultimate load on day one.
That can create opportunities to evaluate phased power delivery.
For example, a campus may begin with an initial block of capacity and increase over time as additional buildings, equipment, or utility infrastructure come online.
A phased strategy may include:
The specific structure depends on the project.
The important point is to align procurement with the actual development schedule.
Where conventional grid capacity is constrained, developers may also need to evaluate other resources that can complement the primary power path.
Those could include:
These options should not be treated as automatic replacements for utility power.
They should be evaluated based on whether they can help improve flexibility, support portions of the load, or reduce dependence on a single energization pathway.
Power procurement should not happen after the project schedule has already been established.
The energy strategy and development strategy need to inform one another.
A strong procurement plan asks:
How much power is needed?
When is it needed?
What infrastructure is required?
What can be secured now?
What needs to be built later?
What supplemental resources may help bridge the gap?
Those questions can be just as important as the price per kilowatt-hour.
AI data center loads can change quickly.
A campus may begin with one phase of development and expand substantially as additional buildings, servers, cooling equipment, and AI computing infrastructure are added.
That makes future electricity capacity one of the most important procurement considerations.
A power strategy that supports the initial phase but leaves no practical path for expansion can create a significant long-term constraint.
Procurement planning should consider more than the first energization milestone.
Developers should evaluate:
The further ahead these requirements are understood, the more opportunity the development team may have to plan around them.
A useful procurement strategy can separate power requirements into phases.
For example:
Phase 1
Power required to open the first portion of the campus.
Phase 2
Additional capacity required as new buildings or computing infrastructure come online.
Long-Term Build-Out
The total electricity requirement if the site reaches its intended maximum scale.
Each stage may have a different power solution.
The first phase might rely primarily on available utility capacity.
The second may require infrastructure expansion.
Long-term build-out could incorporate additional contracted generation, renewable resources, storage, or other energy infrastructure.
Long-term procurement agreements should also reflect uncertainty.
Data center growth may occur faster than expected.
It may occur more slowly.
Technology may change.
AI computing density may increase.
Energy markets may shift.
A contract that provides strong economics but very little flexibility may become restrictive if the project evolves differently than anticipated.
Important questions may include:
AI data centers can operate for decades.
That makes energy procurement a long-term financial decision.
The apparent cost of electricity at the beginning of a project may not reflect the full cost of the power strategy over its life.
A more complete analysis may need to consider:
The basic cost of electricity remains important.
That can include:
The exact structure depends on the power provider and location.
The project may also need to account for the cost of the infrastructure required to deliver power.
That can include:
These costs can materially change the economics of what initially appears to be a low-cost power option.
Long-term contracts can create price certainty, but they can also reduce flexibility.
Important questions may include:
A procurement structure should fit the development plan rather than force the project into an inflexible operating model.
Some procurement strategies provide more price certainty.
Others expose the project to greater market fluctuations.
Neither approach is automatically better.
The appropriate structure depends on:
Renewable energy can be obtained through several different structures.
These may include:
Each structure has different financial characteristics.
The procurement strategy should evaluate renewable energy alongside the overall power portfolio rather than as an isolated sustainability expense.
Battery storage may add value through demand management, renewable integration, or energy flexibility.
But that value must be compared with:
Battery economics are highly project-specific.
An attractive energy rate may come with:
Conversely, a somewhat higher-cost strategy may provide more:
The objective is to understand the total strategic cost of power, not just the headline electricity rate.
Renewable energy strategy should also evolve with the facility.
If the campus doubles its electricity demand, renewable commitments and local generation opportunities may need to increase as well.
That can include:
The procurement portfolio should be capable of growing with the data center.
The first megawatts matter.
So do the next ones.
A strong power procurement strategy creates a roadmap from initial energization through full development.
The goal is not simply to find enough power to open the data center.
It is to create a credible path to powering the campus as it grows.
California does not have a single electricity structure.
A data center project may be located in territory served by:
That means power procurement has to begin with a clear understanding of the local market.
The same procurement strategy that works in one California city may not work the same way somewhere else.
In an investor-owned utility territory, the data center may depend heavily on the utility for transmission, distribution, interconnection, and delivery of electricity.
The procurement strategy may need to account for:
The utility relationship can become one of the most important components of the project.
Municipal power can create a different development environment.
A city or local power provider may have greater influence over electricity procurement, infrastructure investment, and economic-development planning.
That can create a more direct relationship between:
The data center
The power provider
The city
and
The broader community
Potential questions may include:
Municipal power should not automatically be assumed to be simpler.
It should be evaluated as a distinct procurement environment.
In many California communities, electricity procurement may involve a Community Choice Aggregator while the investor-owned utility continues to manage transmission and distribution.
That creates an important distinction.
The organization procuring the electricity may not be the same organization responsible for delivering it to the data center.
A procurement strategy may therefore need to consider:
Understanding who controls which part of the power relationship is essential.
A project in municipal power territory may have different opportunities than one located in a traditional utility market.
A CCA market may create additional renewable procurement options.
A location with substantial commercial and industrial property may support a broader distributed energy strategy.
That is why Solar Harmonics evaluates procurement at the local-market level.
Power procurement begins with understanding who actually controls the power path.
For many AI data center developers, renewable energy commitments are an important part of corporate sustainability strategy.
But renewable procurement should not be planned separately from the facility’s actual power needs.
The strongest approach integrates renewable energy into the overall procurement portfolio.
Some projects may directly develop solar or other renewable generation.
That can include:
Physical generation can create a direct connection between the project and new energy infrastructure.
PPAs can provide a long-term contractual structure for renewable energy.
Depending on the project, a PPA may support:
The agreement should still be evaluated against the data center’s real operating profile and contract risk.
Some utilities and municipal power providers may offer renewable energy options within their existing supply portfolio.
These programs can sometimes provide a simpler procurement path than developing a dedicated project.
Their value depends on:
Renewable procurement does not have to be concentrated in one large generation project.
A distributed strategy can involve solar and storage across multiple local properties.
Potential sites may include:
This approach can create additional local generation while complementing larger procurement arrangements.
Storage can improve the flexibility of renewable generation.
Solar may produce most strongly during the day while the data center continues consuming power overnight.
Battery systems can help shift portions of that energy to different periods.
Storage does not eliminate the need for dependable power supply, but it can make renewable resources more useful within the overall portfolio.
A strong renewable energy strategy should be able to explain:
What renewable resource is being added?
How does it support the data center’s energy requirements?
How does it fit with the broader procurement strategy?
When those answers are clear, renewable energy becomes part of the infrastructure plan rather than a separate marketing commitment.
The goal is not to maximize one type of energy.
It is to create a portfolio that can support:
The strongest renewable energy strategy is the one that makes the overall power procurement strategy stronger.
For an AI data center, power procurement should begin early enough to influence development decisions.
Waiting until design, permitting,g or construction is well underway can reduce the number of practical options available.
The strongest time to evaluate procurement is often during site selection or early development, when the project team still has flexibility around location, phasing, infrastructure, and energy strategy.
Evaluate a potential site not only for land, fiber, zoning, and tax considerations, but also for the quality of its power path.
Important questions may include:
Two sites that look similar from a real estate perspective may have very different energy profiles.
Those differences can affect both project timing and long-term economics.
Utility agreements can shape the project for years.
Before major commitments are made, developers should understand:
The objective is not to avoid utility service.
For most large data centers, the utility or municipal power provider will remain central to the energy strategy.
The objective is to understand the full relationship before the project becomes dependent on it.
Power purchase agreements can provide valuable long-term renewable energy and pricing benefits.
They can also create long-term contractual obligations.
Before entering a PPA, the project should understand:
A renewable energy agreement should support the operating strategy rather than become a constraint if the project changes.
If utility infrastructure is taking longer than the development schedule allows, the procurement strategy should be revisited early.
That may mean evaluating:
Not every alternative will be practical.
The value comes from identifying which options deserve analysis before the project reaches a critical timing constraint.
Power procurement should also be revisited when an existing data center prepares for major growth.
Expansion can change:
The procurement strategy that worked for the original facility may not be sufficient for the next phase.
Starting early is not simply about securing electricity sooner.
It is to preserve strategic flexibility.
The more decisions that remain open, the more opportunities the development team may have to create a stronger long-term power portfolio.
Electricity can influence nearly every major decision surrounding an AI data center.
It can affect:
That makes power procurement one of the most consequential decisions in the development process.
Solar Harmonics helps developers look beyond a single electricity contract and evaluate the larger power environment surrounding the project.
An initial power procurement review can examine:
The best procurement strategy is not necessarily the one with the fewest components.
It is the one that provides the right combination of:
Every project will balance those priorities differently.
The important step is understanding the tradeoffs before major energy commitments are made.
If you are planning, developing, or expanding an AI data center in California, Solar Harmonics can help identify the procurement options surrounding the project and determine where additional energy resources may create value.
Most power procurement strategies focus almost entirely on the electricity delivered directly to the data center.
That is necessary, but it may not capture every available opportunity.
The Solar Harmonics approach also evaluates whether the surrounding community contains energy resources that could become part of the broader strategy.
That may include:
The electricity from every project does not necessarily need to flow directly into the data center.
The larger objective is to understand how additional energy infrastructure can be developed around the project and how that investment may complement the primary procurement strategy.
Warehouses, logistics centers, manufacturing facilities, and other large commercial properties may have significant rooftop or parking-area potential.
These properties can be evaluated for:
A cluster of suitable properties can create a distributed renewable energy portfolio around the data center market.
Cities may also control properties that could be evaluated for energy development.
Potential sites can include:
Where technically and financially viable, these projects may create a stronger connection between data center growth and local energy investment.
Some sites may be more valuable for storage than for solar.
Battery projects can potentially support energy flexibility, local capacity, or renewable integration depending on the specific market and interconnection conditions.
The project may also have access to larger offsite energy opportunities that complement the primary supply contract.
These can be evaluated alongside:
This broader approach matters because data centers are increasingly discussed not only as economic-development projects, but as major new electricity consumers.
Cities and communities may ask:
Where will the electricity come from?
What infrastructure will be required?
Who will pay for it?
What renewable energy is being added?
What benefits will exist locally?
A broader procurement strategy can help make those questions part of the development conversation rather than treating them as separate public-relations issues.
Traditional procurement asks:
Where can we buy the electricity?
The Solar Harmonics approach adds another question:
Where can additional energy resources be developed around the project?
That distinction connects power procurement directly to the Solar Harmonics Framework.
It creates the possibility that a data center’s electricity demand can become a catalyst for additional local generation and storage rather than functioning solely as a new load on the system.
Data center power procurement is the process of securing the electricity and related energy resources required to develop and operate a data center.
For a large AI data center, that can involve much more than opening a utility account.
A complete procurement strategy may include:
The goal is to create a power portfolio that can support the project’s requirements for availability, reliability, timing, cost, sustainability and future growth.
As early as possible.
Ideally, power procurement should begin during site selection or early development.
Electricity availability can influence whether a site is viable at all.
Developers should understand:
Waiting until a site is fully committed can reduce flexibility and make some energy options more difficult to pursue.
AI data centers can be developed faster than the electric infrastructure required to serve them.
That creates a timing mismatch.
A developer may be ready to build, but the utility may still need to complete:
If those improvements take longer than the project schedule, the power timeline can become the development bottleneck.
That is why procurement should evaluate not only where electricity will come from, but when it can realistically become available.
A power purchase agreement, or PPA, is a contract under which a buyer agrees to purchase electricity or energy-related value from a generation project over a defined period.
For a data center, a PPA may be used to support:
PPAs can be associated with resources such as:
The structure of the agreement matters.
Important considerations can include:
A PPA should fit the data center’s actual operating profile rather than be evaluated only as a renewable energy transaction.
No.
Utility service and a PPA serve different roles.
Utility service typically involves the local power provider delivering electricity through the transmission and distribution system.
A PPA is a contractual arrangement tied to a particular energy resource or portfolio.
A data center may use both.
For example, the facility may receive physical electricity through the local utility while also entering long-term renewable energy agreements that support its broader procurement strategy.
The exact structure depends on the market and project.
Yes, if the project is located in territory served by a municipal or locally controlled power provider.
Municipal power can create a different procurement environment than an investor-owned utility.
Potential differences may include:
Because the city and power provider may be more closely connected, municipal power can also create opportunities for a broader discussion about local energy investment.
The structure varies significantly by community, so it should be evaluated on a project-specific basis.
A diversified power portfolio uses several complementary energy resources rather than relying entirely on one source.
A portfolio might include:
Utility or Municipal Power
for primary electricity supply
Long-Term Power Agreements
for pricing and supply certainty
Renewable Energy
to support sustainability objectives
Solar
for physical renewable generation
Battery Storage
for energy flexibility
Distributed Energy
for additional local generation and storage
Future Capacity
for data center expansion
Not every project needs all of these resources.
The objective is to determine which combination best fits the specific site and development plan.
The procurement strategy should consider the full expected campus build-out, not just the first phase.
Developers should understand:
A project that can secure its first 50 MW but has no credible path to the next 100 MW may face a serious long-term constraint.
That is why procurement planning should create an energy roadmap from initial energization through full build-out.
Yes.
Many data center campuses are developed in stages, and the power strategy can be phased as well.
For example, a project might use:
Initial Phase
Available utility or municipal capacity
Expansion Phase
Additional utility infrastructure and contracted power
Later Phases
New generation, renewable procurement, solar, storage or other resources
A phased strategy can align energy investments with the actual development schedule.
The critical issue is making sure later capacity has a credible path before the project becomes dependent on it.
Not automatically.
A data center may operate continuously while renewable resources such as solar and wind vary by time of day and weather.
A project can procure a large amount of renewable energy on an annual basis without physically matching its load every hour.
A more advanced strategy may consider:
The appropriate level of matching depends on the company’s sustainability goals, procurement philosophy and project economics.
At minimum, the project should understand:
The agreement should also be tested against several possible futures.
What happens if the data center opens late?
What happens if it grows faster than expected?
What happens if electricity demand is lower?
What happens if an additional campus phase is added?
Long-term contracts should support those possibilities rather than create unnecessary constraints.
The goal is not simply to obtain the lowest electricity rate.
It is to build a power strategy capable of supporting the project over its entire development life.
That means balancing:
Availability
Time to Power
Reliability
Long-Term Cost
Contract Flexibility
Renewable Energy
Infrastructure
Expansion
Community Considerations
The strongest procurement strategy gives the developer a credible path from the first megawatt to the full campus build-out.
In many California communities, a Community Choice Aggregator, or CCA, procures electricity for customers while the investor-owned utility continues to manage transmission, distribution and delivery.
That means different organizations may control different parts of the power relationship.
A data center procurement strategy may need to understand:
The CCA does not replace the utility’s responsibility for the wires and grid infrastructure.
But it can be an important participant in the energy procurement strategy.
No.
The lowest quoted energy price may not represent the lowest total strategic cost.
A low-cost supply option may still create challenges if it involves:
A somewhat more expensive option may provide advantages such as:
The decision should consider total long-term value, not only the initial price per kilowatt-hour.
A complete analysis may need to consider more than the energy rate.
Potential costs can include:
A strong procurement strategy looks at the combined economics over the expected life of the project.
Yes.
Solar can be incorporated through several different approaches.
These may include:
Solar usually should not be viewed as the only power source for a large AI data center.
It is typically more useful as one component of a broader portfolio that also includes dependable power supply and, in some cases, battery storage.
Its value depends on where and how it is deployed.
Yes.
Battery storage can complement the procurement strategy by giving the project more flexibility in when electricity is stored and used.
Depending on the application, batteries may support:
Battery storage does not create electricity by itself.
Its value depends on what it charges from, how it is operated, and what problem it is intended to solve.
They may reduce or reshape portions of utility demand, but they do not automatically eliminate the need for utility infrastructure.
Large AI data centers can require continuous power at a scale far beyond what most onsite solar systems can produce.
Solar may support portions of daytime load.
Batteries may shift electricity to different periods.
Distributed energy may add generation throughout the surrounding market.
But the data center may still rely heavily on utility or municipal power.
The more accurate objective is often to create greater energy flexibility and diversification, rather than complete utility independence.
Yes.
Distributed energy can become part of the broader procurement and energy strategy.
Potential locations may include:
These projects may include:
Not every distributed project needs to physically deliver electricity directly to the data center to have strategic value.
The larger opportunity is to add new local energy resources around the market where the data center is creating substantial new demand.
The procurement strategy can consider energy investment outside the data center itself.
Potential opportunities may include:
This creates a different community conversation.
Instead of focusing only on how much electricity the data center consumes, stakeholders can also evaluate what additional energy infrastructure may be developed alongside the project.
These opportunities are not guaranteed and must be evaluated technically, financially, and contractually.
But including them in the procurement conversation can reveal options that would otherwise be missed.
No.
Solar Harmonics does not replace the utility, municipal power provider, energy supplier, engineer, attorney, or specialized energy-market advisor.
Our role is to help developers evaluate the broader power strategy surrounding the project.
That can include:
Specialized technical, legal, and contractual work should be performed by the appropriate professionals as the project advances.
The Solar Harmonics Framework expands procurement beyond the question of where the data center buys electricity.
It asks whether the project’s energy demand can also become a catalyst for additional local generation and storage.
That may mean evaluating:
Data Center Demand
Utility or Municipal Power
Renewable Procurement
Solar
Battery Storage
Commercial and Industrial Properties
Municipal Energy Opportunities
Community Participation
The primary power supply remains essential.
The framework adds another layer by looking for energy resources that may exist throughout the market surrounding the data center.
The more project information available, the more focused the initial evaluation can be.
Useful information may include:
Not every item needs to be known before the first conversation.
The purpose of the initial review is often to identify which questions need to be answered next.
Key Points San Jose has spent decades at the center of the digital economy. Now the capital of Silicon Valley is confronting the physical infrastructure
Key Points Pittsburg, California, is emerging as an important test case for one of the biggest questions facing the artificial intelligence economy: How can communities
Key Points Hayward, California, is becoming part of a much larger transformation taking place across the San Francisco Bay Area. Artificial intelligence is driving demand
Key Points The San Francisco Bay Area helped create the technology behind today’s artificial-intelligence revolution. Now the region faces a different challenge: How will we
Key Points Introduction A 500 MW data center is not an ordinary commercial development. When a major AI infrastructure developer approaches a city, the proposal
Key Points The race to build artificial intelligence infrastructure is increasingly becoming a race for electricity. Land can be acquired. Buildings can be constructed. GPUs
