Measurement and Verification (M&V) is the data-driven engineering process that calculates the actual demand reduction delivered by a resource during a grid event. It compares real-time meter data against a Customer Baseline Load (CBL)—a counterfactual estimate of what consumption would have been without intervention—to isolate the precise megawatt impact.
Glossary
Measurement and Verification (M&V)

What is Measurement and Verification (M&V)?
Measurement and Verification (M&V) is the rigorous analytical process of quantifying the actual load reduction delivered by a demand response resource against its statistical baseline to determine financial settlement.
The M&V process ensures settlement integrity by applying statistical adjustments for weather, occupancy, and operational variability. It transforms raw interval data into a verified performance metric, directly feeding the settlement engine to calculate payments or non-performance penalties in ancillary service markets.
Core Methodological Pillars
The rigorous analytical process of quantifying the actual load reduction delivered by a demand response resource against its baseline to determine financial settlement.
Customer Baseline Load (CBL) Calculation
The statistical foundation of M&V that establishes a counterfactual: what would consumption have been without the event? Common methods include:
- High X of Y: Selects the 10 highest usage days from a 45-day window, averaging the top X to capture peak potential
- Regression models: Fit weather-adjusted baselines using temperature, humidity, and day-type variables
- Metering before/metering after: Direct comparison of pre-event and during-event interval data
- Control group methodology: Uses non-participating customers with similar load profiles as a reference Accuracy is measured by Relative Precision and Bias Error, with regulatory bodies like CAISO and PJM mandating specific CBL methodologies for market participation.
Performance Metrics & Settlement Quantification
The translation of raw meter data into financial outcomes through precise calculation of:
- Actual Load Reduction (kW): CBL minus measured event-period consumption, calculated per settlement interval
- Performance Ratio: Actual reduction divided by committed capacity, determining payment multipliers or penalty triggers
- Ancillary Service Scoring: For frequency regulation, metrics include delay time, ramp rate compliance, and precision score against the automatic generation control signal
- Baseline Adjustment Factors: Symmetric additive or multiplicative adjustments applied to CBL for weather anomalies or occupancy changes Settlement engines ingest these metrics to calculate capacity payments, energy payments, and non-performance penalties per market tariff rules.
Statistical Validity & Uncertainty Analysis
Rigorous quantification of confidence in measured savings to satisfy regulatory audit requirements:
- Coefficient of Variation (CV): Standard deviation of savings divided by mean savings; values below 0.3 indicate acceptable precision
- t-statistic testing: Determines if observed load reduction is statistically distinguishable from zero at a 90% or 95% confidence level
- Sampling error propagation: When measuring aggregated portfolios, M&V protocols account for meter accuracy class (typically ±0.2% for revenue-grade meters) and communication dropout imputation
- Persistence validation: Longitudinal analysis confirming that savings endure beyond the initial measurement period without degradation Regulatory frameworks like the International Performance Measurement and Verification Protocol (IPMVP) define four options (A, B, C, D) for varying rigor levels.
Metering Infrastructure & Data Acquisition
The hardware and communication layer that feeds the M&V engine with interval data:
- Revenue-grade meters: ANSI C12.20 Class 0.2 or 0.5 accuracy, providing 15-minute or sub-15-minute interval data via DNP3 or Modbus protocols
- Submetering: Behind-the-meter sensors isolating specific loads (HVAC, lighting, industrial processes) for granular verification
- Phasor measurement units (PMUs): Synchrophasor data at 30-60 samples per second for fast-responding resources providing frequency regulation
- Data validation, editing, and estimation (VEE): Automated cleansing routines that flag and impute missing or anomalous intervals using linear interpolation or same-day analogs
- Time synchronization: GPS-locked timestamps ensuring alignment between utility SCADA and resource meter data to avoid settlement disputes.
Non-Performance Penalty Structures
The enforcement mechanisms ensuring demand response resources deliver committed capacity:
- Availability penalties: Applied when a resource fails to respond to a dispatch signal, typically calculated as a percentage of capacity payment forfeited
- Energy deficiency charges: Levied per kWh of shortfall between committed and delivered reduction, often priced at real-time locational marginal price plus a penalty multiplier
- Consecutive failure rules: Escalating sanctions for repeated non-performance, potentially leading to suspension from market participation after 3-5 consecutive failures
- Make-whole provisions: Requirements for underperforming resources to compensate the market operator for the cost of procuring replacement capacity
- Force majeure exemptions: Documented equipment failures or communication outages that excuse performance, subject to post-event verification.
IPMVP Framework & Regulatory Alignment
The globally recognized standard governing M&V methodology selection:
- Option A: Partially Measured Retrofit Isolation: Combines measured key parameters with stipulated values for non-critical variables; lowest cost but highest uncertainty
- Option B: Retrofit Isolation with All Parameter Measurement: Direct measurement of all variables affecting the isolated system; preferred for demand response due to high accuracy
- Option C: Whole Facility Analysis: Uses utility main meter data with regression models; suitable for whole-building DR programs
- Option D: Calibrated Simulation: Employs building energy models calibrated against actual meter data; used when baseline data is unavailable Alignment with ISO/RTO market manuals (PJM M&V Manual 11, CAISO DR Protocols) ensures settlement compliance and audit defensibility.
Frequently Asked Questions
Clear answers to the most common questions about the rigorous quantification of demand response performance and financial settlement.
Measurement and Verification (M&V) is the rigorous analytical process of quantifying the actual load reduction delivered by a demand response resource against its Customer Baseline Load (CBL) to determine financial settlement. It serves as the impartial accounting engine that bridges physical grid operations and market payments. The process involves collecting interval meter data before, during, and after a demand response event, applying a standardized baseline methodology, and calculating the delta between observed consumption and the counterfactual baseline. Without robust M&V, demand response markets cannot function because there is no trusted mechanism to verify that a promised kilowatt-hour reduction actually occurred. The Settlement Engine relies entirely on M&V outputs to calculate payments and penalties for participants, making it the financial backbone of Virtual Power Plant (VPP) operations and Ancillary Service Markets.
Enabling Efficiency, Speed & Accuracy
Intelligent Analysis, Decision & Execution
We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.
Talk to Us
Search across company data
Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.
Useful when people spend too long searching or get different answers from different systems.

Automate internal workflows
Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.
Useful when repetitive work moves across multiple tools and teams.

Add AI to products and internal tools
Build assistants, guided actions, or decision support into the software your team or customers already use.
Useful when AI needs to be part of the product, not a separate tool.
Related Terms
Measurement and Verification relies on a precise interplay of baseline modeling, settlement systems, and performance metrics to ensure financial integrity in demand response programs.
Customer Baseline Load (CBL)
The statistical counterfactual that estimates what a resource's consumption would have been absent a demand response event. CBL methodologies typically analyze a rolling window of recent, non-event days to establish a reference load shape. Accuracy is paramount, as an inflated baseline directly translates to unearned payments. Common methods include HighXofY, regression models, and matched-day approaches, each with distinct biases depending on weather sensitivity and load volatility.
Settlement Engine
The backend financial system that ingests verified performance data and market rules to calculate payments and penalties. The engine compares metered load reduction against the CBL, applies performance factors, and generates invoices. It must handle complex logic including materiality thresholds, symmetric versus asymmetric settlement, and multi-market stacking rules to ensure regulatory compliance.
Performance Metrics
Quantitative indicators used to evaluate demand response execution. Key metrics include:
- Load Drop: The absolute kW reduction during an event window
- Performance Factor: Ratio of actual to committed reduction
- Availability Score: Uptime percentage of the controllable asset
- Latency: Seconds between dispatch signal and load response These metrics directly determine financial bonuses or non-performance penalties.
Baseline Adjustment
A post-hoc correction applied to the CBL to account for non-routine conditions present during the event day but absent from the baseline window. Adjustments may be triggered by weather deviations, facility occupancy changes, or production schedule shifts. Without proper adjustment, a factory that idled production for reasons unrelated to the DR event could be incorrectly credited with load reduction.
Metering & Telemetry
The hardware and data pipeline infrastructure that captures interval meter data at the required granularity—typically 1-minute to 15-minute resolution. Revenue-grade meters with ANSI C12.20 certification ensure accuracy. Telemetry must be time-synchronized and resilient to communication dropouts, as missing data during an event window can void settlement for that resource.
Statistical Validity Testing
Rigorous analysis to confirm that the observed load reduction is not attributable to random variance. Techniques include t-tests, confidence interval analysis, and coefficient of variation thresholds. Regulators often mandate a minimum statistical significance level before a resource can participate in capacity markets, ensuring that claimed reductions represent genuine controllable flexibility rather than noise.

About the author
Prasad Kumkar
CEO & MD, Inference Systems
Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.
His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.
Partnered with leading AI, data, and software stack.
How We Work
Custom AI workflows for your Business
One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.
01
Review the use case
We understand the task, the users, and where AI can actually help.
Read more02
Pick the right approach
We define what needs search, automation, or product integration.
Read more03
Build the first useful version
We implement the part that proves the value first.
Read more04
Improve from there
We add the checks and visibility needed to keep it useful.
Read moreThe first call is a practical review of your use case and the right next step.
Talk to Us