Data Requirements

Model

The optimization application is the most sensitive to data requirements. To achieve accurate optimization results the best quality data is required. This includes:

Regulating Parameters: The application needs to know the regulating abilities of all the controllable devices: Substation LTCs, Voltage Regulators, Capacitors and Reactors.

Full information including regulation ranges, number of taps, bandwidths, setpoints and so on.

Transformer Impedances: In order to calculate losses correctly an accurate model of the Substation Xfmrs is required.

Conductor Catalogs: Similarly conductor material and construction methods impact losses as power flows through the conductors.

Load Models: How individual loads respond to changes in voltage will affect most of the optimization strategies. An accurate understanding of the relative amount of Constant Power, Constant Current and Constant Impedance is required.

SCADA

SCADA is essential for real-time optimization to function correctly. The optimizer needs to have an accurate understanding of the current network state and what options are available to make adjustments. The required inputs are:

• Device Status (Open/Closed) or On / Off

• Regulation targets

• Tap positions

• Voltages

• Bandwidths (if variable)

Some method of adjusting the device's behavior is necessary. Therefore each device needs to have at least some combination of the following NMS outbound controls:

• Device Status

• Tap Position (Up/Down or set)

• Regulation target

• On / Off thresholds

• Bandwidth

Bellwether Meters

The Volt/Var optimization engine can leverage bellwether meters to provide feedback that allows the ADMS to gauge how effective the previously executed adjustments have been and whether more or less aggressive action is required to achieve the desired optimization objective.

The term bellwether is used to indicate that the ADMS does not need comprehensive voltage measurements at all points along a feeder for IVVC/CVR to work. Good results can be achieved with a small number of meters. So for example, if AMI data is available at every customer meter, the ADMS would only want to consume a small fraction of those meters on a fast (near real-time) channel.

Additionally, the term "Meter" is used in a generic sense of any device capable of measuring voltage. It doesn't have to be an AMI meter on a customer's premise. It can be voltage measurements on the primary circuit located near (or at) a capacitor bank for example.

For the purpose of guiding and directing Volt/Var optimizations, the ADMS only leverages voltage measurements. Other measurements (for example, Amps, kW or kVAR) can be used to sure up the powerflow / state estimation calculations. Typically these types of measurements are most suitable gathered from devices on the primary feeder (for example, reclosers). To date it has been impractical and cost prohibitive to gather loading measurements, in near real-time from all customer meters. It is also questionable whether this exercise would yield significant operational benefit over existing methods to estimate load from recloser data.

A summary of what can be measured and it's effects on various ADMS functionality is given below:

1. For IVVC the only measurement that is processed is voltage. That is used as feedback into the CVR module. It is used to confirm (or refute) that expected changes have occurred and to set limit constraints if the change in voltage is more (or less) than the expected change from the power flow solutions. Essentially it used to refine the expected behavior as determined by the model, with the actual behavior seen in the field.

2. As mentioned above the ADMS can utilize a voltage measurement anywhere along the feeder. That includes:

a. Primary voltage measurements at any existing location where there are measurements. For example a pole-top RTU at a recloser; at this location we might already be monitoring things like: Recloser open/close status, recloser mode (Auto-reclose or switch only mode), fault targets, current flow). The ADMS can utilize primary side voltages from this location too.

b. Secondary voltage measurements either at a service transformer or at a meter. For example the ADMS could consume voltage measurements from an AMI system if they can be obtained in near real-time. If the ADMS is to use measurements at secondary locations, then the model needs to have some basic understanding about the configuration of the secondary network at that location. If the secondary network is not modeled then at the very least an estimate is needed of the percentage voltage drop from the service xfmr to the meter location.

3. Where present, measurements would need update within the frequency at which CVR calculations are run. The ADMS optimization objects are configurable (on a substation transformer basis) to run in Fast mode and Slow mode. Slow mode means once and hour at the top of the hour. Fast mode means on the back of FLM cycles, which are typically expected to be once every 15 minutes (but can be configured differently).

In this situation, suppose there is a set of CVR feeders in fast mode. CVR will run every 15 minutes. Measurement data from the bellwether meters needs to be produced and consumed faster than that. It is recommend that at least two updates between CVR cycles, to confirm that adjustments have had expected results. If CVR is being run on 15 minute cycles it is recommended that bellwether data would be configured on 5 minute cycles.

4. Size and locating of metering:

a. Anecdotal evidence from utilities suggests around 10 measurement points along a feeder provides a good trade-off between expense and accuracy.

b. Measurements located at controllable devices (Substation LTC buses, cap banks and regulators) would be a significant benefit.

c. Measurements at locations identified to be the weakest (lowest) / strongest (highest) voltage levels.