VoltaGrid
Owns and operates modular gas-fired generation fleets under multi-year contracted power agreements (power-as-a-service) for data centers, oilfield electric-frac, and industrial customers; vertically integrating manufacturing via the Propell acquisition
Only the May 2026 round is a confirmed priced equity event with disclosed structure; the 2025 ~$4B figure is a press-reported mark whose round structure was not publicly detailed, and the Feb 2026 point reflects a reported process valuation rather than a closed transaction. Private-company marks; not audited.
Earnings, margins, COGS & capex
Private LLC - no public filings. What is disclosed: >$1.9B capital deployed, >900 employees (pre-Propell), a 4.3+ GW fully contracted deployment plan through 2028 within a ~7.5 GW order book to 2030, and Bloomberg-reported EBITDA reaching ~$1.1B by 2028 versus roughly $200M in 2024 (implied from the >5x statement). The model is contracted power sales (multi-year agreements with data-center and oilfield customers), so revenue visibility is high but so is upfront capex - hence the $5.0B Nov 2025 debt package and the $1B May 2026 equity round.
Revenue trend
Margins
up
n/a
COGS structure
Not disclosed. Structurally: natural gas fuel (firm supply via Energy Transfer's 140,000-mile pipeline and storage network for the Oracle fleet), engine procurement (INNIO Jenbacher reciprocating engines packaged into 25 MW units), balance-of-plant (ABB automation and electrical systems), field operations and maintenance labor, and logistics. The Propell acquisition internalizes QPac manufacturing to cut unit cost and lead time.
Capex
The dominant cash use: building and deploying the QPac generation fleet. The $2.0B notes + $3.0B ABL (Nov 2025) fund the 4.3+ GW contracted plan through 2028; VoltaGrid is also building two additional automated Propell plants in Granbury, TX targeting ~300 MW/month of manufacturing capacity across reciprocating engines and turbines.
Latest earnings
n/a
Press-reported trajectory: ~$1.1B annual EBITDA by 2028 from contracted deployments, >5x the 2024 level (Bloomberg, Feb 2026)
- Order book
- ~7.5 GW through 2030 (May 2026)
- Fully contracted plan
- 4.3+ GW through 2028 (Nov 2025)
- Last round
- $1B ($775M primary + $225M secondary), Blackstone Tactical Opportunities + Halliburton, May 2026, >$10B valuation
- Debt package
- $2.0B senior secured 2L notes due 2030 + $3.0B ABL (closed Nov 10, 2025)
- QPac unit
- 25 MW modular packs (INNIO Jenbacher); nodes combine to deliver up to 200 MW of prime power under a minor-source air permit
- Oracle contract economics
- >$1B/year for a single OpenAI-linked Texas site (Bloomberg-reported, Feb 2026)
Growth drivers
- AI data-center power shortage — multi-year grid interconnection queues push hyperscalers to behind-the-meter gas (Oracle 2.3 GW Texas fleet, Oct 2025; Vantage 1+ GW partnership, Feb 2025)
- Order book expansion from a 4.3+ GW contracted plan (through 2028, Nov 2025) to a ~7.5 GW order book (through 2030, May 2026) in roughly six months
- Vertical integration via Propell Energy Technology — ~300 MW/month manufacturing target de-risks delivery of the backlog
- Halliburton channel — electric-frac power (Chesapeake/Expand since 2021, Diamondback since Dec 2024) plus a 400 MW Eastern Hemisphere data-center manufacturing commitment for 2028 delivery
- Firm fuel supply partnership with Energy Transfer's 140,000-mile pipeline network
Bull & bear
VoltaGrid owns the scarcest asset in the AI buildout - deployable megawatts on a months-not-years clock - with a contracted multi-gigawatt backlog, strategic capital from Blackstone and Halliburton, and now its own factory. If it converts the ~7.5 GW order book, ~$1.1B of 2028 EBITDA makes the $10B mark look like a reasonable entry, not a peak.
- Speed-to-power is the binding constraint on AI capacity; VoltaGrid's modular QPac is among the only proven at-scale answers (Oracle chose it for 2.3 GW)
- Backlog is contracted, not speculative - revenue visibility through 2030 with hyperscaler-grade counterparties, including Bloomberg-reported >$1B/year from Oracle for a single site
- Propell acquisition attacks the #1 execution bottleneck (engine packaging capacity, ~300 MW/month target at Granbury)
- Halliburton is both investor and channel: oilfield e-frac base load plus a 400 MW international data-center commitment
- $5B debt + $1B equity means the buildout is funded without near-term dilution pressure
- At ~$1.1B 2028E EBITDA, a >$10B valuation plus ~$2B+ of debt implies roughly 10-12x forward EV/EBITDA - defensible against public speed-to-power comps (e.g. Solaris Energy Infrastructure) if the contracted backlog converts on schedule
A leveraged, single-technology bet priced for flawless execution: $5B of debt capacity against undisclosed current financials, extreme customer concentration in Oracle, and a bridge-power window that grid buildout, turbines, and eventually SMRs will close. If AI capex decelerates or deployments slip, the equity sits behind $2B of second-lien notes.
- Valuation rests on a projected >5x EBITDA ramp to 2028; today's actuals (~$200M 2024 EBITDA per Bloomberg's implied figure) support a fraction of $10B
- Oracle concentration (~2.3 GW) means one customer's AI-capex decision materially moves the backlog; the site's OpenAI linkage adds second-order exposure
- Second-lien notes due 2030 + ABL create refinancing risk exactly when the bridge-power window may be narrowing
- Engine supply is concentrated; INNIO dependence (even with Propell packaging) leaves prime-mover risk
- Competitors are flooding in: PROPWR, Solaris, Liberty Energy, Aggreko, Cat/Cummins dealers, plus utilities accelerating interconnects - contract pricing on renewals could compress sharply
- Gas-fired data-center power faces permitting and community opposition; minor-source permitting strategies are drawing scrutiny
- No public disclosure: investors underwrite management-provided projections without audited visibility
What it is worth
Last priced round + forward-EBITDA cross-check (no public price; private-round mark)
$4-6B
AI capex slows or Oracle deployments slip, engine/permitting bottlenecks delay the ramp, and leverage (2L notes due 2030) compresses the equity back toward the earlier press-reported ~$4B mark (WSJ, 2025)
~$10B
the May 2026 round mark holds; backlog converts broadly on schedule toward ~$1.1B 2028 EBITDA
$15B+
order book grows past 7.5 GW with new hyperscaler logos, the Propell ramp hits ~300 MW/month, and an IPO prices at a premium to Solaris-type comps on contracted growth
The May 2026 strategic round (Blackstone Tactical Opportunities + Halliburton, $775M primary + $225M secondary) valued the company at more than $10B, consistent with Bloomberg's Feb 2026 report of an IPO or sale being weighed at a valuation topping $10B. Cross-check: against Bloomberg-reported ~$1.1B 2028E EBITDA, $10B+ equity plus roughly $2B+ of debt implies ~10-12x 2028E EV/EBITDA - demanding but defensible versus public speed-to-power comps if the contracted backlog converts on schedule. The mark depends on management-provided projections; no audited financials exist publicly. Not financial advice.
SWOT
Strengths
- Largest portable natural-gas reciprocating-engine power provider in the US (company-disclosed), with a proven fleet operating record since the Aug 2021 Chesapeake Marcellus e-frac deployment
- Contracted multi-gigawatt backlog with blue-chip counterparties (Oracle, Vantage, Halliburton, Diamondback)
- Speed-to-power advantage — modular 25 MW QPac units deploy in months versus multi-year grid interconnection timelines
- Deep-pocketed strategic backers (Blackstone, Halliburton) plus $5B institutional debt access (Goldman Sachs-led notes; JPMorgan Chase-agented ABL)
- Vertical integration of manufacturing (Propell) secures engine-packaging capacity in a supply-constrained market
- Ultra-low criteria-emissions design (up to ~98% criteria-pollutant reduction claimed) eases air permitting versus diesel
Weaknesses
- No public financials — revenue, margins, and leverage metrics are undisclosed, so the $10B+ mark rests on projections
- Heavy leverage early: $2B second-lien notes plus a $3B ABL against a still-scaling EBITDA base
- Customer concentration: Oracle alone is ~2.3 GW of the order book
- Execution risk on a >5x EBITDA ramp in roughly four years — manufacturing, permitting, and field-ops must all scale simultaneously
- Single-technology dependence on gas reciprocating engines (INNIO Jenbacher) amid an industry-wide engine/turbine shortage
Opportunities
- US data-center power demand projected to require tens of GW of new capacity this decade; bridge-power and prime-power both addressable
- IPO or strategic sale (explored per Bloomberg, Feb 2026 — Blackstone and BlackRock among parties that discussed buying or investing) could crystallize value and fund the next growth tranche
- International expansion via the Halliburton 400 MW Eastern Hemisphere commitment
- Adjacent markets: utility peaking support, mining, industrial microgrids, grid-emergency capacity
- Propell's turbine capability broadens the platform beyond reciprocating engines
Threats
- Grid interconnection catch-up or nuclear/SMR maturation could compress the bridge-power window post-2030
- Well-capitalized competitors scaling the same playbook (Solaris Energy Infrastructure, ProPetro PROPWR, Caterpillar, Cummins, GE Vernova turbines, Bloom Energy fuel cells, Liberty Energy - which signed its own 1 GW Vantage deal)
- Natural-gas price spikes or Permian/Waha basis disruptions raising delivered power cost
- Environmental/permitting backlash against gas-fired data-center power (minor-source permitting strategies could face challenges)
- An AI-capex slowdown would hit exactly the hyperscaler demand underwriting the backlog
- Rising rates or credit tightening on a leveraged, capex-heavy balance sheet
Moats, dependencies & bottlenecks
Moats
logistics, field ops at GW scale) Proven since the Aug 2021 Chesapeake e-frac deployment; few rivals have operated comparable reciprocating fleets at this scale
Vantage, Halliburton) contract terms through ~2030 Switching a live behind-the-meter power provider mid-contract is costly and operationally risky
~300 MW/month target) Real in a supply-constrained market, but capacity advantages erode as engine makers expand
minor-source strategy) regulatory-dependent Up to 200 MW prime under minor-source permits is a genuine time advantage; also a regulatory risk if rules tighten
Scale purchasing of engines (INNIO's largest power-delivery order to date, plus a 1.5 GW follow-on) Priority allocation in an engine shortage; not durable once supply normalizes
Dependencies
Sole disclosed prime-mover supplier for QPac; 92 x 25 MW packs for the Oracle fleet (INNIO's largest power-delivery order) plus a 1.5 GW / 300-engine follow-on order (Feb 2026, delivery by 2028)
2.3 GW Texas AI data-center fleet (Oct 2025) - largest single customer commitment; Bloomberg reports >$1B/year for one OpenAI-linked site
Firm gas supply from its 140,000-mile pipeline and storage network underpins the Oracle deployment
partner/investor/customer E-frac channel since 2021, co-investor in the $1B round, 400 MW Eastern Hemisphere data-center commitment for 2028 delivery
Lead of the May 2026 $1B strategic round via Tactical Opportunities (~90% of the round per Bloomberg)
JPMorgan Chase-agented $3B ABL) $5B package funds the buildout; notes mature 2030
Backlog conversion assumes sustained data-center construction through 2030
Advantages
- Months-not-years power delivery versus grid interconnection queues
- Fleet mobility and modularity — redeployable 25 MW units, prime-to-peak flexibility, ultra-fast transient response without battery reliance
- Ultra-low criteria emissions (up to ~98% criteria-pollutant reduction claimed) versus diesel gensets eases siting near communities
- Cross-cycle demand base: oilfield electric-frac (Halliburton/Diamondback/Expand) hedges data-center timing
- Strategic shareholder alignment (Halliburton channel, Blackstone capital) plus vertically integrated manufacturing
Weaknesses
- Opaque financials - no audited public statements to verify the EBITDA ramp
- High leverage relative to the disclosed earnings base
- Customer and technology concentration (Oracle; INNIO reciprocating engines)
- Gas-price and gas-logistics exposure in delivered-power economics
- Valuation embeds flawless multi-GW execution through 2028
Bottlenecks
- Engine and generation-equipment supply (industry-wide reciprocating engine and turbine shortage; the Propell deal is the mitigation)
- Manufacturing/packaging throughput until the two new Granbury plants reach the ~300 MW/month target
- Air permitting cadence per site (the minor-source strategy must clear state regulators at every deployment)
- Skilled field-operations labor to run distributed GW-scale fleets (750+ new Texas jobs planned around the Oracle deployment)
- Firm gas transport capacity at data-center sites outside the Energy Transfer footprint
Top signals & trends
Top signals
An IPO would force disclosure and could re-rate the sector; a sale crystallizes the mark
The key execution de-risker for the 7.5 GW order book
Largest customer exposure; any slip flows straight to backlog conversion
Diversification away from Oracle concentration is the single best fundamental signal
Sets the read-across for any IPO valuation
Regulatory tightening would slow the speed advantage
Trends
The core demand driver; multi-year queues push hyperscalers behind the meter
Reciprocating engines and turbines are the only at-scale near-term answer; VoltaGrid is a category leader
Raises barriers for new entrants but constrains VoltaGrid's own deployment pace - hence Propell
VoltaGrid's original market keeps growing with Permian simul-frac adoption
Long-run substitute that could shorten contract renewals and terminal value
Minor-source permitting strategies are drawing regulatory attention across states
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Advent International-owned) Reciprocating gas engines; QPac co-developer; 92 x 25 MW packs for the Oracle fleet (largest order to date) plus a 1.5 GW / 300-engine follow-on order (Feb 2026)
Propell Energy Technology (acquisition signed May 2026, closing mid-2026) QPac manufacturing, integration, R&D, and after-sales in reciprocating engines and turbines; Granbury, TX plants expanding toward ~300 MW/month
Automation and electrical systems inside the QPac platform
Firm natural-gas supply and transport from its 140,000-mile network for the Oracle Texas fleet
2.3 GW modular gas fleet for OCI AI data centers across Texas (announced Oct 15, 2025)
DigitalBridge-backed) 1+ GW prime-power partnership across North America (announced Feb 2025)
E-frac power customer/channel plus a 400 MW Eastern Hemisphere data-center manufacturing commitment for 2028 delivery
Four electric simul-frac fleets (~200 MW) powered in the Permian with Halliburton ZEUS (Dec 2024 agreement)
First operator on VoltaGrid e-frac power (Marcellus, Aug 2021, with Halliburton)
Public pure-play mobile gas-turbine power for data centers and oilfield; closest listed comp
Oilfield services player scaling a dedicated mobile gas power fleet for data centers and industry
Oilfield power specialist that signed its own 1 GW on-site power deal with Vantage Data Centers
Global incumbent in temporary/mobile power, expanding into data-center bridge power
Cat/Solar Turbines gensets and turbines sold or rented through dealers into the same behind-the-meter demand
Gas reciprocating gensets for prime and standby data-center power
Aeroderivative and heavy-duty gas turbines for larger on-site data-center power blocks
Fuel-cell alternative for behind-the-meter data-center power with faster permitting in some states
Linear generator technology targeting the same flexible on-site power niche
Builds gas-powered AI data centers directly - vertically integrated demand-side competitor
Competes in the electric-frac power segment, VoltaGrid's original market