GridSight® NSE · Network State Estimation.
Answers.
What is GridSight® Network State Estimation (NSE)?
GridSight® NSE infers the complete electrical state of a network - voltage at every bus, current through every branch, power at every point - from the limited measurements that actually exist. Since sensors are concentrated at major substations, most of a distribution network is electrically unobserved. GridSight® NSE makes the invisible visible.
How is GridSight® NSE different from traditional state estimation?
Classical Weighted Least Squares (WLS) estimation needs more measurements than unknowns - a condition rarely met in distribution networks - and either fails or produces unreliable estimates when data is sparse. GridSight® NSE instead computes a feasibility envelope: the full set of operating states simultaneously consistent with network physics (Kirchhoff's laws) and operational constraints. It delivers the range of possible states, the most probable point, and deterministic compliance guarantees - if a safety limit lies entirely outside the envelope, it cannot be violated under any feasible condition.
What data does GridSight® NSE require?
The network model (bus definitions, branch parameters, transformer data), available real-time measurements from connected substations, and pseudo-measurements such as smart-meter data, load forecasts, and distributed-energy-resource output estimates. GridSight® NSE is specifically designed to work when measurement coverage is minimal.
How does GridSight® NSE detect energy theft (non-technical losses)?
NSE compares the electrical state it infers at each point in the network against the energy actually metered or billed downstream. When modelled power injection in a zone doesn't reconcile with measured feeder or substation totals, GridSight® NSE flags the discrepancy - so inspection crews are directed only to zones with a real energy balance gap, instead of sweeping entire feeders.
What can utilities achieve with GridSight® NSE?
With GridSight® NSE, utilities can find energy that was used but not billed, spot technical and non-technical losses, manage networks with very little visibility, and see exactly how much a new sensor would improve that visibility before installing it.