The Energy Stack.
All 13 Diagrams.

Master Energy Stack

The complete LegacyGrid energy operating system

An overview of all five layers working together — Tesla Megapack BESS, waterless cooling, heat recovery, workload scheduling, and community resilience. LegacyGrid's preferred storage hardware is the Tesla Megapack (~3.9 MWh per unit). For a 60MW Phase 1 facility with 4-hour backup, that's approximately 240 MWh — or ~62 Megapack units. This is the system, not just a battery.

LegacyGrid AI — Energy Stack Overview

Peak Shaving

Cut demand charges at peak hours

Tesla Megapack discharges during peak demand windows to reduce the school's highest-demand billing tier. In Texas, demand charges can represent 30–50% of a commercial electricity bill. At ~3.9 MWh per Megapack unit, even a modest 10-unit deployment (~39 MWh) creates meaningful demand charge reduction.

LegacyGrid AI — Energy Stack · 01 — Peak Shaving

Battery absorbs demand spikes — utility sees a flat, predictable load.

Grid Protection

Ride through grid instability without interruption

Tesla Megapack acts as a buffer between the utility grid and the data center — absorbing frequency deviations, voltage sags, and momentary outages before they reach sensitive compute equipment. Megapack's sub-100ms response time is faster than any diesel generator handoff.

LegacyGrid AI — Energy Stack · 02 — Grid Protection

BESS isolates the data center from grid instability — voltage sags, frequency deviations, and surges.

Outage Ride-Through

Keep compute running when the grid goes down

Tesla Megapack provides minutes-to-hours of backup power — enough to ride through most outages or allow an orderly generator handoff. A 60MW Phase 1 facility targeting 4-hour backup requires ~240 MWh of storage, or approximately 62 Megapack units. Critical for data center uptime guarantees and SLA commitments.

LegacyGrid AI — Energy Stack · 03 — Outage Ride-Through

Keep compute running when the grid goes down.

Energy Arbitrage

Charge cheap, discharge expensive

Tesla Megapack charges during low-price overnight hours (ERCOT off-peak can drop to $20–30/MWh) and discharges during high-price peak periods (ERCOT peaks can reach $150–200/MWh+). That price spread, multiplied across 240 MWh of capacity, turns storage into a genuine revenue asset — not just an insurance policy.

LegacyGrid AI — Energy Stack · 04 — Energy Arbitrage

Charge cheap. Discharge expensive. Turn storage into a revenue asset.

Community Resilience Reserve

30% of BESS ring-fenced for the community

LegacyGrid's model requires that a portion of Tesla Megapack capacity be reserved for campus and community use — not just data center backup. In a 62-unit deployment (~240 MWh), roughly 18–20 units (~72 MWh) would be ring-fenced for the university and surrounding community. This is a guardrail written into the deal structure, not a marketing claim.

LegacyGrid AI — Energy Stack · 05 — Community Resilience Reserve

A dedicated partition of BESS storage is reserved for the surrounding community during grid outages.

AI Workload Scheduling

Software turns storage into an intelligent operating system

The intelligence layer reads battery state, energy price, and grid stress signals in real time — dispatching urgent jobs immediately, shifting flexible workloads to cheap-energy windows, and curtailing load during grid stress events.

LegacyGrid AI — Energy Stack · 06 — AI Workload Scheduling

Software turns storage into an intelligent operating system.

Student Training + Replication

Universities become builders, not just land hosts

Live infrastructure becomes a training platform. Students in data center operations, BESS management, cybersecurity, and fiber networking get paid internships and certifications — then replicate the model across the Texas HBCU network.

LegacyGrid AI — Energy Stack · 07 — Student Training + Replication

Universities become builders, not just land hosts.

Capture Data Center Heat

95–99% of electrical input becomes heat

Nearly all server power becomes heat. Liquid cooling and hot aisle containment capture it at the source — before it becomes waste. This is the first step in the heat recovery chain.

LegacyGrid AI — Energy Stack · 08 — Capture Data Center Heat

95–99% of electrical input becomes heat. Capture it at the source.

Reuse Heat First

Direct reuse is the practical first win

Send recovered heat to campus buildings, hot water preheat, and controlled agriculture before attempting electricity conversion. Direct reuse has the highest efficiency and lowest capital cost.

LegacyGrid AI — Energy Stack · 09 — Reuse Heat First

Direct reuse is the practical first win. Highest efficiency, lowest capital cost.

Steam / Electricity Recovery

Secondary value loop — when conditions are right

ORC turbines and heat pumps can convert recovered heat to electricity — but only when coolant temperatures exceed 60°C, capital economics are favorable, and heat volume is sufficient. Not every site will qualify.

LegacyGrid AI — Energy Stack · 10 — Steam / Electricity Recovery

Secondary value loop — when conditions are right.

Water Recovery / Condensation

Zero water in. Recovered water out.

Serverfarm already runs waterless cooling — zero evaporative consumption. LegacyGrid adds a heat-driven condensation recovery layer on top: warm exhaust air hits a cooling surface, humidity condenses, and recovered water is reused for campus irrigation, cooling support, and agriculture. The data center stops being a heat waste site and becomes a water recovery asset.

LegacyGrid AI — Energy Stack · 11 — Water Recovery / Condensation

Zero water in. Recovered water out.

Serverfarm already runs waterless cooling — zero evaporative consumption. LegacyGrid adds a heat-driven condensation recovery layer on top.

Integrated Energy Ecosystem

Not a single battery idea — a complete campus operating system

All layers working together: utility grid + PPA, Tesla Megapack energy yard (~62 units / 240 MWh for Phase 1), AI data center, heat recovery, waterless cooling, student ops lab, and community resilience reserve. This is the LegacyGrid model. Source: Tesla Megapack specifications (tesla.com/megapack/design).

LegacyGrid AI — Master View · 12 — Integrated Energy Ecosystem

Not a single battery idea — a complete campus operating system.

Intelligent Developments (EDC-4)

Neighborhood-scale distributed AI compute — the fourth layer

EDC-4 brings xFRA-compatible distributed AI inference nodes to Texas HBCUs, K-12 schools, and new construction — each paired with a site-level BESS system that buffers the AI compute load from ERCOT. At scale across the Texas HBCU corridor, the collective BESS capacity becomes a Virtual Power Plant earning ancillary service revenue. This is the layer that connects LegacyGrid's campus-scale model to the broader distributed compute ecosystem that SPAN and NVIDIA are building.