This node is provided to enable modellers to input a time series of demands. It is potentially useful when calibrating a Source model where a staged approach is used, for example where it is desired to calibrate storage behaviour using recorded historical outflows and inflows. It can be used when modelling both unregulated and regulated river systems.
Scale
Implemented at the site scale.
Principal developer
eWater CRC
Scientific Provenance
Providing a time series of demands as input data is a basic, and well established, approach in river system and storage modelling.
Version
Source version 2.17.
Dependencies
A Water User node is required by Source to access the time series demand node.
Structure & processes
Time series of demands can be input:
- via a data file;
- from an existing scenario; or
- via an expression defined by the modeller.
The Time Series Demand node provides demand values to the Water User node at every model time step. When modelling a regulated system and:
- The time series of demands have been input via a data file, the Water User node supplies the Time Series Demand node with a value of maximum travel time, tmax (defined below). At a given model time step, the demand value extracted from the input data file is the one for tmax time steps in advance of the current time step.
- Demands come from an existing scenario. In this case the procedure operates in the same way as when the demands have been input via a data file. The information loaded from an existing scenario must have the same number of time steps as used in modelling the current scenario.
- Demands are calculated by evaluating the expression defined by the modeller, no adjustment for travel time can be made as expressions can only be evaluated based on current information (ie they cannot "look ahead"). In this case the demand value can only be the one that is applicable to the current model time step (equivalent to tmax=0).
Where tmax = the travel time from the upstream storage having greatest travel time between it and this Time Series Demand node, expressed in numbers of model time steps.
In the first two cases, the extracted data value is used to update an array of demand values for the current time step and tmax time steps into the future, and this array is returned to the Water User node. In the third case, only a single value is returned to the Water User node.
When modelling an unregulated system, the demand value returned to the Water User node is the one that is applicable to the current model time step, regardless of whether it comes from an input data file, from a scenario, or from evaluating an expression.
More information on how this value is further processed is available in the Water User section of this Scientific Reference Guide.
Input data
When demands are entered in a time series data file or demands come from an existing scenario, there must be one value per model time step with no missing values. Units must be specified. Information on other input data is given in Parameters or settings, below. Details are provided in the Source User Guide.
Parameters or settings
Model parameters are summarised in Table 30.
Note that when the return flow is being calculated no mass balance check is carried out. While there should be no potential for a mass balance problem when the return flow is evaluated as a percentage or proportion of the water supplied, when either a fixed volume or a volume defined by an expression is used there is the potential to cause mass balance problems if an inappropriate value of the fixed volume or an inappropriate expression is used.
Units allowable are:
- Megalitres per day
- Gigalitres per day
- Megalitres per year
- Cubic metres per second
- Litres per second
- Kilolitres per day
- Kilolitres per year
Output data
The Time Series Demand node returns an array of demand values to the Water User node, or a single demand value for the current model time step, as described above.
Table 30. Model parameters
...
Parameter
...
Description
...
Units
...
Default
...
Range
...
Travel time (only when modelling a regulated system)
...
Expected time for water to travel from nominated storage to the demand node
...
Days
...
0
...
Governed by maximum travel time in the system being modelled
...
Demand data
...
A file of demand values
...
See list below
...
0.0
...
Positive number
...
Information from an existing scenario
...
An expression defined by the modeller
...
Return Flow (see note below)
...
Percentage of volume of water supplied (which may differ from demand)
...
Percentage or proportion
...
A fixed volume
...
Units can be selected
...
A volume defined using an expression
...
A time-series demand model (shown in Figure 1) allows you to extract water directly from a storage and identify the amount of water required to satisfy a demand at each time-step. The model does not have to be directly downstream of the storage, the supply point will order water from the storage to meet demands.
If a confluence node is connected between a storage node and a water user demand node, the flow pathway link in the confluence must be set to Regulated for the storage to order water to satisfy the demand downstream. If it is left unregulated, water will not be ordered from the storage and the supply point will pull water from the river at the point in the network that it sits to meet the demand. if the demand is higher than the volume in the reach, then the supply point will extract everything.
If there is no storage in the network above the supply point and water user node, then the volume is extracted from the river at the point where the supply point sits in the network.
Figure 1. Water user node (Time series demand model)
Time-series demands can be specified using either a single value, a time-series data file (Table 1 shows the format of a file) or a function using the Function Editor.
| Info | ||
|---|---|---|
| ||
| Note: A time series demand model is used to ensure that there is no discrepancy in orders between leap and non-leap years when specifying a monthly pattern demand model. Use a pattern variable with daily units (rather than monthly) to allow Source to recognise the number of days in February. As an example, consider the data input for a water user demand model which has been assigned a function ($Function1), which in turn is a variable of type time series ($ts). Figure 2 shows the time series variable (notice that Results Units are specified in megalitres per day) and Figure 3 depicts the input function assigned to the demand model. |
Figure 2. Time series demand model, Time series variable
Figure 3. Time series demand model, Input function
Refer to Return flows for details on configuring this parameter.
Table 1. Water User node (Time Series Demand, data file format)
Row | Column (comma-separated) | |
|---|---|---|
1 | 2 | |
1 | date | value |
Where:
date is the date of observation in dd/mm/yyyy format (eg. 31-12-2000)
value is the observed value (eg. 2600).