FINESSE – hydraulic simulation, optimal pump scheduling and leakage management package

From FINESSE web page:

Network Simulation
Provides steady state hydraulic simulation, and ג€œextended-timeג€ reservoir dynamics (e.g. over a time horizon). The calculation engine is the GINAS software, which since the 1980s, has been applied by many water companies to many networks.

On-line Network Simulation
Is a special configuration of the Network Simulation function for on-line application. It performs hydraulic simulation using the most recently measured state of the network as the initial condition. The Network Simulator is configured to automatically acquire data including reservoir levels, flows and pressures. It then performs a simulation over a predefined time horizon.

Network Scheduling
Calculates least cost control schedules and reservoir profiles. Pumps, valves and water works outputs are scheduled to minimise the total network energy and water production costs, typically over a time horizon of one or more days. The schedules conform to practical operating constraints defined by the user. Recent advances use non-linear hydraulic models to calculate least cost schedules for large networks. Such models are accurate over wide ranges of operating conditions and are very important in the analysis of operating constraints. The Scheduler uses the network simulation model directly.The modelling and calculation engines are based on the established GAMS and CONOPT software, which are used in a number of different market sectors for optimisation. GAMS makes it relatively easy to reformulate optimisation problems to take into account any new cost-functions, system components and constraints required by users.

Pressure Control
Is a tool to calculate optimal pressure profiles directly from simulation models of a network. The profiles are calculated to minimise leakage whilst maintaining specified head constraints at each node of the network. The pressure control module can be applied to large pressure management areas of any configuration (single feed, multi feed etc.).

Demand Prediction
Predicts water demand patterns and volumes from historical flow time-series. Typically, 6 weeks of historical data are required to set-up the prediction model, which is based on categorisation of demand patterns and a triple exponential smoothing algorithm. Thereafter, the package predicts future patterns of demand (e.g. for 1 to 7 days at 15 minute intervals) from recently acquired time-series. The Predictor can be configured to acquire data on-line for a number of zones and then predict demands patterns for each zone. Alternatively, it can also be used to analyse historical flow time-series off-line.

Model Simplification
Optimisation is computationally intensive and model simplification may be required for large network models. The Model Simplifier automatically calculates a simplified model directly from a detailed one. The simplified model is accurate and contains all key components and any critical pressure nodes chosen by the user. Simplified models usually accurate to within 2% of the original model. Simplified models can be used in either simulation or optimisation and they can be recalculated from the detailed model at any time if there are any changes to the network. The Simplifier is based on a Gaussian elimination procedure.

SCADA Gateway
Provides an interface to supervisory control and data acquisition (SCADA) systems for on-line applications. In the case studies it has been customised to interface with two SCADA systems, one provided by ASPENTECH, the other is home made developed by a water company. The user specifies the data servers and the mnemonics for the on-line variables in a simple look-up table. The table also maps the variables to model parameters. FINESSE can also be interfaced to other SCADA and telemetry systems.

More information can be found on the Water Software Systems web-site.