Definition
TCDR measures the rate at which a distribution system digital twin's calibration accuracy is degrading over time, expressed as the change in root mean square error per unit time.
The Measurement Gap
Measurement Gap and Decision Supported
The Invisible Failure Mode: A water utility calibrates its distribution system hydraulic model as part of a master plan update. At calibration, the model matches field pressure measurements within ±2.1 PSI across 45 monitoring points: well within the industry-accepted threshold. Three years later, the utility uses the same model for emergency response planning, pressure zone optimization, and capital project hydraulic analysis. Nobody has rechecked calibration since the master plan. The system has changed: 14 miles of new development pipe, 23 valve replacements, a new booster station, and gradual increases in C-factor degradation in aging cast iron mains. The twin's RMSE has drifted from 2.1 PSI to 6.8 PSI, but no one has measured this because the model still runs and produces reasonable-looking outputs. During a transmission main break, the emergency response team uses the twin to identify isolation valve sequences and predict pressure impacts, and the model's recommendations are wrong enough to cause a low-pressure event in a hospital service area.
What Existing Metrics Miss:
Digital twin accuracy is typically checked at calibration time and then assumed stable until the next master plan update, a cycle that often spans 5–10 years. No standard metric tracks the rate at which calibration is degrading between major updates. Without TCDR, utilities discover calibration problems reactively when the model fails during operational use.
The Decision TCDR Enables:
Is the digital twin's calibration degrading fast enough to require recalibration before the next planned update, and which areas of the distribution system are experiencing the fastest divergence? TCDR provides a condition-based recalibration trigger rather than relying on a fixed time interval.
Cross-References:
TCDR is the temporal complement to Digital Twin Fidelity Index (DTFI, GMI-WT-007) from the maintenance KPI set. DTFI measures the twin's accuracy at a point in time; TCDR measures the rate at which that accuracy is changing. A utility needs both: DTFI tells you the current gap, and TCDR tells you how fast that gap is growing. DFHIS (GMI-WT-025) feeds into TCDR because demand forecast errors that create systematic pressure deviations propagate through the hydraulic model and accelerate calibration drift. Twin Scenario Coverage Ratio (TSCR, GMI-WT-027) adds a third dimension: even a well-calibrated twin may not cover the operational scenarios that matter most.
Formula
Plain language: At each monitoring point with actual pressure/flow data, compute the model's prediction error (predicted vs. actual). Track this error over time. The divergence rate is the slope of the prediction error over rolling quarters.
Notation:
TCDR = Delta(RMSE) / Delta(t)
Where:
- RMSE = root mean square error between twin-predicted and actual measurements across all validated monitoring points for a given quarter
- Delta(t) = time between measurement periods (quarters)
- RMSE = sqrt(Sum((Predicted_i - Actual_i)^2) / N_points)
Units: RMSE units (PSI or GPM) per quarter. Positive TCDR = divergence increasing (model aging). Zero = stable accuracy. Negative = improving (unusual without recalibration).
Proposed formula, pending validation pilots.