If you are evaluating a microgrid for your facility, the technology conversation is the easy part. The harder question is whether the financial model holds up under scrutiny: your scrutiny, your board's scrutiny, and the assumptions that tend to fall apart between the proposal and the signed contract. This is a framework for thinking through that model clearly and getting to a number you can actually defend. 

Start With What You're Already Paying 

The strongest microgrid business cases are built on costs you are already incurring, not projections about what you might save. 

Pull your last 12 months of utility bills. For most commercial and industrial facilities, two line items will immediately stand out: energy charges and demand charges. Demand charges (the fee your utility levies based on your highest 15-minute consumption window in a given month) typically represent 30 to 50 percent of the total bill. That number is the foundation of your model. It is this recurring line item that a well-designed microgrid attacks most directly. 

The second number to pull is your outage history. How many hours of unplanned downtime has your facility experienced in the past two to three years? What did each event cost in lost production, spoiled inventory, SLA penalties, emergency response, or recovery time? If you operate a data center, a cold storage facility, a manufacturing plant, or any asset where downtime has a hard dollar cost, that number belongs in your model. Resiliency value is not a soft benefit. It is a quantifiable reduction in operational risk. 

Those two figures, demand charge and downtime cost, are the anchors that everything else builds on. 

The Five Value Streams That Drive the Return 

A complete microgrid ROI model captures five distinct sources of value. Leaving any one out understates the return and creates a gap that will surface during due diligence. 

Demand charge reduction is typically the largest line item. Intelligent battery storage dispatch shaves your peak demand window consistently, reducing the demand charge that resets every month. The annual savings here is calculable from your utility rate structure and your load profile and does not require assumptions about future energy prices. 

Energy arbitrage captures the spread between off-peak and on-peak pricing. In time-of-use markets, the difference between what you pay overnight versus peak afternoon hours is significant and recurring. A storage system that charges during low-price windows and discharges during high-price windows captures that spread automatically. The value depends on your market, your rate structure, and your storage capacity, all of which are knowable before you commit. 

Resiliency value converts your downtime exposure into a financial figure. If your facility costs $50,000 per hour of downtime and you have experienced an average of six hours of outage exposure per year, that is $300,000 in annual risk. A microgrid with islanding capability will not eliminate outage exposure entirely, but even a conservative estimate puts the reduction at 30 percent or more, which is a significant dollar figure for any facility with real downtime costs. 

Federal and state incentives directly reduce your effective capital outlay. The federal Investment Tax Credit for standalone battery storage is currently 30 percent of system cost, with additional credits available for domestically manufactured equipment and certain qualifying locations. State-level incentives vary but can be substantial. The correct way to model this is against net capital cost, not gross. A system that looks marginal at full cost often looks compelling once incentives are applied. 

Grid services revenue applies in markets where your storage can participate in demand response, frequency regulation, or capacity programs. This is market-dependent and should be modeled conservatively. Treat it as upside rather than base case unless you have confirmed program eligibility and contracted rates. 

How the Model Comes Together 

Stack those five value streams against your net capital cost (equipment, installation, commissioning, and interconnection fees, less incentives), and you have the inputs for a straightforward NPV calculation. 

A few things are worth building in explicitly rather than leaving as footnotes: 

Degradation. Battery capacity declines at roughly 2 to 3 percent per year. Your Year 10 savings are meaningfully lower than your Year 1 savings. A model that holds Year 1 performance constant over a 15-year project life overstates value and will not survive scrutiny. Build the degradation curve in from the start. 

Interconnection timeline. In many markets, grid interconnection queues have lengthened significantly. If your model assumes a six-month interconnection process and the actual timeline runs to 18 months, your cash flow timeline shifts, and your payback period moves with it. Get a real interconnection estimate before finalizing the model. 

Curtailment. If your solar generation capacity exceeds your facility's load and storage capacity to absorb it, you are effectively wasting generation during peak production hours. Your model should reflect actual utilization, not theoretical maximum output. 

Financing structure. If you are evaluating a Power Purchase Agreement rather than direct ownership, model the PPA rate against your current and projected utility rate over the contract term. The question is not whether the upfront cost is zero, it is whether the locked-in PPA rate delivers savings relative to what you would otherwise pay, compounded over 15 years of utility rate escalation. 

What a Defensible Model Looks Like 

A model you can stand behind has three versions: a base case, an upside case, and a downside case. The downside case uses conservative savings assumptions, higher capital cost, and a longer interconnection timeline. If the project still delivers an acceptable return under the downside case, that is the number worth leading with, because it is the number that holds up when someone stress-tests your assumptions. 

Simple payback period and NPV at your hurdle rate are the outputs that matter most for internal approval. IRR is useful for comparing against other capital projects. All three should come from the same model, not from separate calculations that can diverge under questioning. 

The technology behind a microgrid is proven and widely deployed. The financial case, built correctly on site-specific data rather than industry benchmarks, is the thing that turns an interesting option into a committed project. 

NextNRG works with utilities, logistics operators, healthcare systems, and commercial facilities to develop site-specific microgrid models built on actual load data, real interconnection timelines, and verified performance figures. Visit nextnrg.com to learn more.