Diversity calculation (dialog)
The Diversity calculation dialog: time shift, minimum load at peak, detailed settings, target diversity curve, plus chart and consumer time shifts.
VICUS Districts Page 3 of 3 in this chapter
Overview
The Diversity calculation dialog bundles the settings, calculation and results of diversity. It is opened via the Calculation settings… button in the Diversity and part load tab (with diversity activated). The dialog works on a copy of the project; only Apply to project transfers the results.
When the dialog is opened, previously saved results are discarded - the calculation must therefore be started again with Calculate each time it is opened.
Settings
The upper part contains the most important settings:
| Option | Default | Range | Meaning |
|---|---|---|---|
| Maximum time shift: | 6.00 h | 0 - 24 h | Maximum shift of each demand profile forward or backward. Larger values offer more scope for peak-load reduction but become less realistic; typical values are 3 - 6 h. Values above 10 h are possible, but their physical significance then decreases. A value of 0 deactivates the shift |
| Minimum load at peak: | 30 % | 0 - 100 % | At the time of the network peak load, every consumer draws at least this share of its own maximum power; consumers below it are shifted again after the optimization. This ensures that all consumers are supplied in the steady-state calculation. A value of 0 % disables the constraint, see Minimum load at the network peak |
| Target diversity curve: | Winter et al. | – | Via the Edit … button the target curve is inspected and changed in the Edit diversity dialog (see below) |
The collapsed Detailed settings section contains further options:
| Option | Default | Range | Meaning |
|---|---|---|---|
| Calculation accuracy: | Balanced | Fast / Balanced / Thorough | Trade-off between speed and result quality. Fast searches around one peak day only and with shorter search runs. Balanced and Thorough search three peak days; Thorough uses twice as many restarts and deeper search runs, which can improve the results for large networks. Balanced is recommended for most cases |
| Time step for demand profiles: | 20 min | > 1 - 1440 min | Temporal resolution with which the profiles are discretized. Smaller values are more accurate but slow down the calculation; the value must be an integer > 1 and must divide 1440 (= minutes per day) without a remainder (e.g. 10, 15, 20, 30, 60). Typical: 15 - 30 min |
| Minimum diversity: | 0.65 | 0 - 1 | Lower limit of the target diversity curve: wherever the curve drops below it, the optimization targets the limit instead. This keeps the target realistic for routes with many consumers. A value of 0 uses the curve unchanged, see Lower limit of the target curve |
If one of these values (except the target curve) is changed while results already exist, the query Discard diversity results? appears. With Yes the results are discarded, with Cancel the input is reset.
Practical tip
The Maximum time shift is the most effective lever on the result - and at the same time the most easily overdone. Larger values spread the peaks out more strongly and lower the diversity factor, but assume that demands can really be shifted by several hours. Stay in the plausible range of 3-6 h; values above 10 h yield nicer factors but lose their physical significance. For the Calculation accuracy, Balanced is the right choice in most networks.
Running the calculation
The Calculate button starts the optimization. A progress bar appears, the calculation can be stopped with Abort. If the time step is not a divisor of 1440, the dialog reports Invalid time step and aborts.
After a successful calculation the dialog reports Optimization complete with the number of calculated consumers and - provided routes were evaluated - the average deviation from the target. In addition, the message states how many consumers were shifted after the optimization so that they draw at least the Minimum load at peak at the network peak, and for how many consumers this minimum load could not be reached within the maximum time shift (the closest possible shift is used there).
Results
The results appear in two tabs that are only active once valid results exist:
- Diversity plot - shows the Target curve and the Calculated diversity per route over the Number of consumers. The target curve is shown as the optimization aims at it, i.e. limited from below by the Minimum diversity. Only routes with at least three downstream consumers are shown as points.
- Time shifts per consumer - table with the columns Consumer and Time shift [h].
Applying the results
The Apply to project button transfers the settings and results into the project; it is only operable with valid results. Cancel discards all changes in the dialog. Changes to the network, demand or target curve invalidate stored results - they must then be recalculated before being applied.
Editing the target diversity curve
The Edit … button next to Target diversity curve: opens the Edit diversity dialog. It shows the target curve on the left as a chart and in the center as a lookup table with the columns Number of buildings [-] and Diversity [-].
| Element | Meaning |
|---|---|
| Lookup table | Each row is a curve point and can be edited directly. The Number of buildings must be an integer > 0 and must run in ascending order; the Diversity must lie in the range > 0 and ≤ 1 |
| Set default function | After a query, resets the curve to the default function according to Winter et al. (Euroheat & Power, 2001); support points up to 295 buildings |
To the right of the table there are four icon buttons: add a curve point (5 buildings further, 90 % of the last diversity value), remove selected points (the curve always keeps at least two points), copy all points tab-separated to the clipboard, and paste two-column data from the clipboard (after a query).
With OK the changed curve is adopted, with Cancel it is discarded. For individual pipes the curve can additionally be overridden by a custom diversity.