As electrification and decarbonization efforts accelerate, utilities and policymakers are deploying an ever-growing toolkit of flexible load solutions: from time-of-use (TOU) rates and demand response to managed EV charging, solar PV, battery storage, and electrified heating. Each intervention offers potential benefits: lower emissions, grid optimization, and customer savings. But what happens when they intersect, overlap, or collide?

Increasingly, the value and effectiveness of one load-modifying intervention can be amplified—or undermined—by another. We’re seeing this play out in the real world. Our recent paper explored several empirical case studies where interactive effects across flexible load programs revealed unexpected synergies, conflicts, and blind spots.
 

The problem: (Un)intended consequences in a multi-DER world

Imagine layering a demand response event with pre-cooling over a household already responding to TOU pricing on a day when the EV is set to charge, solar production is peaking, and heat pumps are running full tilt. Each of these signals might be “smart” on its own. But without coordination, the results can be chaotic: snapback effects, diminished savings, customer confusion, or even reduced participation.

We’ve entered an era where distributed energy resources (DERs) don’t operate in silos. Interactions are dynamic, context-dependent, and often invisible until the effects emerge. Without the right frameworks, utilities risk undercutting the very value they’re trying to unlock.
 

Use Case 1

The (un)intended energy benefits of residential smart thermostat demand response program

Smart Thermostat Demand Response Programs can deliver energy savings as a result of multi-hour load curtailment. However, pre-conditioning and snapback can erode achievable energy savings. In fact, events with aggressive pre-conditioning strategies are more likely to result in negative energy savings (i.e., increase in energy consumption) following demand response events. Event duration also, not surprisingly impacts achievable energy savings.

Our analysis of over two dozen events dispatched between 2019 and 2022 in the Midwest across three device types reveals the following:

• Average energy savings of 0.8 kWh per device per event day or 2% of the event day HVAC energy consumption
• Aggressive pre-conditioning strategies depress energy savings, reaching as much as 4.46 kWh or a 21% increase in daily consumption, and are a threat to customer experience, and can lead to increased bills
• Without the aggressive pre-conditioning strategies, energy savings reach 1.22 kWh per device or 4% of the event day HVAC energy consumption
• Carefully curating event durations and pre-conditioning strategies can allow DR programs to deliver additional energy conservation benefits

Residential Demand Response programs offer opportunities for additional thermostat optimization on non-event days to harvest additional energy savings. Our research for Ameren Missouri’s Residential DR program shows the following:

• Energy optimization on non-event days results in 1.57 kWh in daily energy savings or an 8% reduction in daily HVAC load
• Combined with event day energy savings, energy savings harvested as part of non-event day optimization result in over 235 kWh in energy savings, which translates to approximately $32 in bill savings per participating customer

Use Case 2

The (un)intended impacts of EV managed charging interventions

Managing EV load through pricing signals embedded in the Time-of-Use rates, as well as through managed charging programs, is growing in popularity. A managed charging pilot in the Pacific Northwest demonstrates that multiple EV load management strategies can be successfully deployed with a thoughtful program. The PGE Pilot enrolled customers into one of two load management groups, depending on whether customers were also on a whole-home TOU rate. The Pilot also randomly assigned customers to an unmanaged control group, which provides a baseline to assess the impacts of managed charging, the TOU rate, and the stacking of both interventions. Our assessment of the pilot’s interventions revealed the following:

• When the EV charging load is unmanaged, the TOU rate is effective at shifting the charging load from peak hours of 5:00 p.m. to 9:00 p.m. to off-peak hours.
• Managed charging interventions, however, are also highly effective at shaping participants’ charging load both during the hours coincident with the TOU peak as well as during the hours following the peak.
• Applying both managed charging and TOU interventions targeting the same set of hours may not deliver enough incremental benefit. That said, when staggered in time, these two interventions can target different hours of the day, beneficially shaping the overall charging load.

Use Case 3

Energy conservation and renewable adoption

• Customers with solar achieve deeper energy savings than those without
• Solar adoption is disproportionately lower in low-income homes and disadvantaged communities
• Increased adoption of energy efficiency combined with solar will likely make additional solar generation capacity available

Use Case 4

Building electrification and customer bills

Customer experiences and perceived benefits from home electrification updates are meaningful drivers of building electrification progress

• Customer concerns about bill increases and upfront costs are key barriers to adoption
• Impacts of electrification on customer bills vary by region, with the cost of electricity as compared to fossil fuels being one of the key drivers
• Impacts of electrification on customer bills vary by customer—those with solar and air conditioning are more likely to experience bill savings than those without
• Time-varying rates are a grid resource—not always a bill-saving resource—no consistent relationship between flat vs. time-based rate designs and customer

The solution: FLEX—A framework for integration and insight

To address these complexities, we developed a decision-making approach rooted in systems thinking, adaptive management, and complexity science to guide stakeholders in managing the integration of load-modifying interventions with greater intention, flexibility, and awareness of interactive effects: the FLEX Framework.

FLEX stands for:

• Formulate: Define the solution’s intended outcomes and anticipate its ripple effects on customers, the grid, and the broader ecosystem.
• Link: Examine how the intervention interacts with existing programs or technologies, assessing synergies, conflicts, and future evolution.
• Engage: Involve diverse stakeholders to capture operational, behavioral, and equity insights before
• Experiment: Pilot and iterate with a learning mindset, incorporating real-world data to refine design and implementation.

Monitoring and evaluation are integral, woven into the process to ensure programs adapt to evolving technologies and shifting customer needs.

Grounding the framework

Our FLEX Framework is an adaptable, iterative approach for managing load-modifying solutions within the evolving, multi-DER landscape. FLEX provides a method to approach, structure, and operate to navigate the increasing complexity of DERs. It aids organizations in addressing the intertwined technical, behavioral, and market-driven dynamics of today’s grid while positioning themselves to adapt to future developments.

The framework is grounded in three powerful decision-making and complexity-management theories:

• The Cynefin Framework: A sense-making model that differentiates between complicated and complex environments, guiding decision-making amid uncertainty.
• Adaptive Management Theory: An iterative, evidence-driven approach originally developed for natural resource management, emphasizing learning through practice.
• Systems Thinking: A holistic framework for understanding interdependencies and feedback loops within complex systems.

These theories collectively shape a flexible structure that enables utilities, regulators, and solution providers to intentionally design, deploy, and evaluate flexible load programs, considering both intended and unintended outcomes.

Foundation principles of the FLEX framework

The FLEX Framework is built on six core principles:

1. Acknowledge the interconnectedness and interdependence of the various load-modifying solutions.
2. Confirm the inherent uncertainty and gaps in knowledge, especially in rapidly changing technological and policy environments.
3. Recognize the continuous and rapid change that the grid, technologies, and customer behaviors are undergoing.
4. Encourage flexible, feedback-informed strategies that adapt over time and support iterative experimentation and refinement.
5. Emphasize stakeholder engagement to incorporate diverse perspectives, surface tensions, and improve coordination.
6. Prioritize experimentation and data-driven decision-making, allowing programs to evolve through learning, not just planning.

Why it matters for evaluation and planning

From an evaluation, measurement, and verification (EM&V) standpoint, overlapping load interventions introduce new layers of complexity. Is that midday peak reduction due to rate responsiveness, a smart thermostat’s automation, or managed EV charging? Untangling these threads requires advanced planning, creative experimental designs, and ongoing feedback loops.

Leveraging the FLEX Framework, evaluators and program designers can

• Avoid misattribution of savings,
• Detect unintended consequences earlier,
• Design with customer experience in mind, and
• Strengthen business cases for future investments. 
   

Where we go from here

As the grid modernizes, the conversation must shift from evaluating “point solutions” in isolation to understanding and managing portfolios of interactive interventions. It’s not just about which programs “work”—it’s about how they work together, and for whom.

To navigate this complexity effectively, we need to integrate experimentation, coordination, and systems thinking into our planning and evaluation from the outset.