The Dispatchability Mirage: Demand Response vs. Curtailment

Hero image for The Dispatchability Mirage: Demand Response vs. Curtailment

If you have spent any time in a control center or reviewing interconnection studies, you know the feeling: the grid is saturated, the price signals are screaming, and the dispatch software is making decisions that seem to ignore the physics of the distribution feeder. We love to talk about the “flexibility” of the grid, but when we strip away the marketing, we are left with two distinct, often misunderstood levers: Demand Response (DR) and Curtailment.

Confusing the two is a rookie mistake that leads to poor procurement decisions and, eventually, a phone call from the regional transmission operator (RTO) asking why your site’s telemetry looks like a heartbeat monitor during a frequency event.

The Problem Nobody Talks About

The industry often treats DR and curtailment as interchangeable tools for balancing supply and demand. They are not. One is a contractual agreement involving human behavior or automated load shedding; the other is a brute-force physical limitation imposed on generation assets.

Consider a recent incident at a mid-sized industrial facility operating behind the meter (BTM) with significant solar PV and a battery energy storage system (BESS). The site’s automation system was programmed to prioritize “curtailment” of the solar array when the BESS hit state-of-charge (SOC) limits, rather than initiating a formal DR event to shed facility load. Because the curtailment signal was sent to the inverter communication bus without a corresponding check on the facility’s internal load profile, the site inadvertently triggered an under-voltage condition on the local bus when the solar dropped off too aggressively while the BESS was already in a high-impedance state. The result? A site-wide trip that cascaded into a local utility distribution fault.

This is the reality of demand-response-vs-demand-side-management in practice: if your control logic doesn’t distinguish between reducing load and killing generation, you are essentially flying blind.

Technical Deep-Dive

To understand the difference, we have to look at the power flow equations at the point of interconnection (POI).

Demand Response is fundamentally a modification of the load profile. It is a change in the $P$ and $Q$ consumed by the end-user. Whether it is an interruptible tariff or an automated response to a price signal, DR relies on the site’s ability to reduce its draw from the grid without compromising its internal operational integrity. From the grid operator’s perspective, DR is a “negative load.”

Curtailment, conversely, is the intentional reduction of output from a generation source—usually renewable—below what it is physically capable of producing at that moment. This is done to prevent transmission congestion or to maintain grid stability when the supply exceeds the local load and export capacity.

The distinction is critical for your SCADA architecture. DR requires a communication path to the load controllers (HVAC, pumps, industrial processes), often involving latency-sensitive protocols. Curtailment requires a direct, high-priority write command to the inverter’s active power setpoint. If you use a slow, polling-based protocol like Modbus RTU to manage curtailment, you may find your response times are insufficient to meet the stringent requirements of modern grid codes, potentially leading to non-compliance penalties.

Implementation Guide

When implementing these systems, the engineering objective is to move from reactive manual interventions to automated, telemetry-driven responses.

  1. Telemetry Fidelity: You cannot manage what you cannot see. Ensure your BTM assets are reporting at a high enough frequency to catch transient spikes. If your SCADA update rate is slower than the inverter’s response time to a curtailment command, you will experience oscillation.
  2. Protocol Selection: Use robust, deterministic protocols. While legacy systems still rely on DNP3 or Modbus, modern grid-edge integration should prioritize protocols that support explicit control feedback loops.
  3. Control Hierarchy: Your local controller must have a prioritized list of actions.
    • Level 1: Curtail generation (if the inverter allows for fast, stable power setpoint adjustments).
    • Level 2: Discharge BESS to absorb excess generation or meet load.
    • Level 3: Demand Response (shed non-critical loads).
    • Level 4: Physical disconnect (the last resort, usually involving protective relaying).

If you are designing the control logic, ensure your algorithms account for the ramp rates of your assets. A solar inverter can curtail almost instantaneously, but industrial loads often have thermal or mechanical inertia that prevents them from shedding immediately. Mismatching these ramp rates is the fastest way to destabilize your local bus.

Failure Modes and How to Avoid Them

The most common failure mode is the “communication handshake failure.” In a DR event, the site controller sends a command to the load, but the load controller is in local mode, not remote. The command is ignored, the site fails to shed, and you are in breach of your DR contract.

Another frequent failure is the “setpoint hunting” loop. When an inverter is told to curtail, it may reduce output, but if the grid voltage is volatile, the inverter’s internal voltage-var function may fight the curtailment command. You end up with the inverter oscillating between curtailment and voltage regulation, which puts unnecessary thermal stress on the power electronics.

To avoid these:

  • Hardwire your overrides: Ensure that critical safety or grid-compliance commands can bypass the primary automation network if the network fails.
  • Validate the feedback: Never assume a command was executed. Use real-time power metering at the POI to verify that the load or generation has actually changed.
  • Test the logic in simulation: Use hardware-in-the-loop (HIL) testing to verify how your controllers handle communication loss during a dispatch event.

When NOT to Use This Approach

Do not rely on DR for safety-critical grid stability. DR is an economic and operational tool, not a replacement for protective relaying. If your facility is relying on DR to stay within its transformer capacity during peak hours, you have a design flaw. You should have sized your infrastructure to handle the peak load, or installed sufficient BESS to shave the peak. Using DR as a crutch for undersized infrastructure is a liability waiting to happen.

Similarly, do not use curtailment as a long-term solution for transmission constraints. If you are constantly curtailing your generation, your interconnection agreement is likely flawed, or your site location is suboptimal. Curtailment should be an edge-case tool for extreme grid conditions, not a standard operating procedure for daily energy management.

Conclusion

Demand response and curtailment are powerful, but they require a rigorous engineering approach. They are not magic buttons to solve capacity issues. They are control inputs that require high-fidelity telemetry, robust communication, and a clear understanding of the physical constraints of your equipment.

If you are a procurement professional, stop asking for “grid-ready” solutions and start asking for the specific latency specs of the control loop, the protocol support, and the failure-mode behavior of the controller when it loses the master signal. If the vendor can’t explain how their system handles a communication timeout, they are selling you a headache.

*This article is intended for informational purposes only for experienced electrical engineers and equipment procurement professionals. All specific technical parameters, protocol compliance thresholds, and performance specifications mentioned must be independently verified against the applicable standard revision, equipment datasheet, and site-specific engineering studies before any design, procurement, or operational decision is made. GridHacker and its authors accept no liability for misapplication of the content herein.*

Hero image: A sleek white robotic vacuum cleaner is parked at its self-emptying and charging station in a modern, minimalist room with light hardwood floors and simple decor.. Generated via GridHacker Engine.

Related Articles