Closing Stores in Texas

Retail closes stores continuously. Leases expire, formats change, a market gets overbuilt, a chain rationalizes after an acquisition. The process is well practiced: real estate negotiates the exit, construction manages the fixture removal and the restoration obligation, IT pulls the network gear, and finance books the charge. Every one of those steps has an owner and a date.

The electricity account does not. It sits between real estate, which stops thinking about the site the day the lease terminates, and accounts payable, which has no way of knowing the site is gone because the invoice keeps arriving and keeps looking normal. In ERCOT, where the account is held with a Retail Electric Provider rather than the wires company, that gap is wider still, because closing a store requires a transaction that only the REP can submit and that nobody in the closure workflow has been told to request.

The result shows up in every multi-site portfolio we review in Texas: active ESI IDs at premises the business vacated, quietly billing a customer charge and a minimum demand every month, sometimes for years.

The Move-Out Is a Transaction, Not a Phone Call

Every metered premise in the ERCOT competitive market carries an ESI ID — an Electric Service Identifier that names the service location itself, not the tenant. The ESI ID persists across occupants. What changes is which REP is serving it and in whose name.

Ending service is a market transaction: you request a move-out from your REP, the REP submits it, and the transmission and distribution utility — Oncor, CenterPoint, AEP Texas or TNMP, depending on where the store sits — schedules a final meter read for the requested date. Three things follow from that mechanic:

• Only the account holder can start it. Handing the keys to the landlord does not end your service obligation. If nobody submits the move-out, the ESI ID remains in your name and the meter keeps being read.
• The date is scheduled, not retroactive. Move-outs are effective on a scheduled read. Requesting one in November for a store that closed in July does not usually reverse the intervening four months.
• A dark store is not a closed account. Turning off the breakers reduces consumption to near zero. It does not remove the fixed monthly charges, and it does not remove billing demand where a ratchet applies.

The fix is procedural and cheap: add one line to the store closure checklist that names the responsible party and the ESI ID, and require the move-out confirmation number before the closure is marked complete.

Why a Dark Store Still Has a Demand Bill

Delivery charges on a Texas commercial service are dominated by demand, not consumption. Many commercial delivery rates set billing demand as the greater of the current month's measured demand or a stated percentage of the highest demand recorded in a preceding period, commonly the prior eleven or twelve months. That provision exists so the wires company recovers the cost of capacity it built to serve your peak, and it does exactly what it was designed to do after you leave.

For a Texas retailer this lands badly on the calendar. A store's annual peak is almost always set in the summer, when the rooftop units run continuously through a July or August afternoon. A closure in the fall therefore inherits a ratchet that was set at the hottest moment of the year and will keep setting billing demand into the following spring. On a mid-box format with substantial refrigeration or HVAC, that is not a trivial monthly charge, and it is being paid on an empty building.

The lesson is not that you should close stores in January. It is that the delivery-side cost of holding an unclosed account is much larger than the consumption on it suggests, so the value of submitting the move-out promptly is larger than anyone assumes when they look at a bill showing almost no kWh. Our primer on understanding demand charges covers the ratchet mechanics in general terms.

Termination Liability: The Formula, Not the Fee

If the store is inside a fixed-price supply agreement with volume commitments, closing it raises a second question: what do you owe the supplier for power you contracted to buy and will not consume?

Most Texas commercial agreements answer this with a mark-to-market formula rather than a flat fee. The supplier is deemed to resell your unconsumed volume into the forward market and bills you the shortfall if the current forward price is below your contract rate. The consequences of that structure are worth internalizing before you negotiate a closure schedule:

• The liability is directional. If forward prices have risen above your contract rate since you signed, the supplier is made whole or better by reselling, and the liability may be nominal. If prices have fallen, it is real money.
• It is priced at a moment. Two closures three weeks apart in a moving market can produce materially different bills for identical stores.
• Bandwidth breaches can trigger it without a formal termination. Many agreements permit consumption to vary within a band — often expressed as a percentage above and below the contracted annual volume — and price the excursion outside that band at market. A store going to zero can push the portfolio outside the band even if nobody terminated anything.

The provisions that decide all of this are the same ones we walk through in the energy contract clauses that cost you money. They are negotiated at signing and unnegotiable at closure.

The Better Answer for a Chain: Add and Drop, Not Terminate

A retailer with dozens or hundreds of Texas locations has a structural advantage that a single-site business does not. If the locations sit under one master agreement rather than a stack of individual contracts, a closure becomes a drop within a portfolio rather than a termination of a contract — and the volume can often be reallocated to remaining stores, to new openings, or to the relocation site that replaced the one you closed.

That flexibility has to be written in. The language to insist on covers four things: the right to add premises at the contract rate or at a defined pricing mechanic, the right to drop premises without termination liability up to a stated share of portfolio volume, the treatment of a dropped site's volume against the bandwidth calculation, and whether a relocation is treated as an add plus a drop or as a continuation. A chain with a known closure cadence should be negotiating a drop allowance sized to that cadence as a matter of course. The portfolio mechanics are covered further in multi-site energy procurement.

The Retail-Specific Traps

Beyond the general mechanics, a few things recur specifically in retail decommissioning:

• The sign and parking-lot lighting account. Pylon signs, parking-lot poles and security lighting are frequently on a separate unmetered or small-metered service from the store itself. Closing the store account leaves this one open. It is small, it is durable, and it is the single most common phantom account in retail.
• The landlord handoff that never happened. In many leases the service is supposed to revert to the landlord at surrender. If the landlord does not submit their own move-in and you do not submit a move-out, the market has no instruction and service stays where it is. Never assume the counterparty acted.
• Construction and demolition meters. A temporary service set for a remodel or a demolition is a separate ESI ID and gets forgotten with unusual reliability, because the contractor who ordered it is gone before the final bill arrives.
• Refrigeration left running. In grocery and convenience formats, cases are frequently kept energized after closing while inventory is redistributed. That is a legitimate decision. It should be a decision, with an end date, rather than a discovery.
• Deposits. Commercial deposits held by a REP are refundable at final billing. They are also routinely left unclaimed, because the entity they were paid by has stopped watching the account.

The Sweep: Reconciling the Store List to the Meter List

Every retailer should be running one reconciliation annually, and it takes an afternoon rather than a project. Pull the list of active ESI IDs and utility accounts being paid, pull the list of open store locations from real estate, and compare them. The rows in the first list that do not appear in the second are your findings.

Three qualifiers make the exercise honest. First, expect legitimate mismatches — a distribution center, an office, a sign account at a store that is genuinely open under a different address string. Second, obtain the account history under a Letter of Authorization scoped to bill history only, so the exercise cannot quietly become a supplier switch. Third, act on the findings inside the back-billing window, because the recovery is capped by tariff regardless of how long the error ran.

The larger point is that decommissioning is not an event that happens once. In a chain of any size it is a continuous process, and the energy work belongs inside the closure checklist rather than inside an annual cleanup. The cleanup finds what the checklist failed to prevent. Both are worth doing; only one of them is free.