Calculating the date for ‘Day Zero’ at any moment in time is tricky; it involves a number of factors which can vary from day to day. This results in the date ‘jumping’ about from week to week and can be confusing and disconcerting. We asked permission from DWS to publish the weekly graphs that they use to monitor the system drawdown, which may help a little to manage members’ fears and expectations.
This graph uses red dots to plot the ACTUAL water in the 6 major dams supply the CoCT (Theewatersdkloof, Voelvlei, Berg River Dam, Wemmershoek and the two Steenbras Dams) against PROJECTED levels that are the result of modelling the system. The projected storage assuming low inflows like last year are shown using the dark black dashed line and represent the MINIMUM TARGET. it is important that the storage RED DOTS remain above the BLACK DASHED LINE. That means we avoid dropping the levels of the dams below 13.5%, which gives us a narrow safety margin of about 5% before the water in the dams becomes unusable.
Keep RED above BLACK and we’re on TRACK…
If you look closely at the red dots you will see that around the 29th January they were beginning to diverge below the minimum target. However, subsequently due to the reduction in the combined agricultural and urban demands and the donation of some water by the Groenland Water User Association from Eikenhof Dam the red dots are starting to converge back towards and even beyond the target. This is in essence why “Day Zero” was shifted out into July.
It is conceivable that a wetter winter will be experienced than last year and the various coloured lines represent the probability of these scenarios. For instance, there is about a 50% chance that the storage will follow the green line and reach 700 million m3 providing the planned demand savings are achieved. Similally, there is a 10% chance of the dams filling and less than 2% chance that similar inflows to last year will be experienced.
The system was modelled taking the following variables into account:
c) Urban potable usage from the dams supplying CoCT
d) Agricultural usage from the dams supplying CoCT
It did NOT account for:
a) Augmentation from CoCT boreholes and desalination plants
b) Augmentation from Eikenhof dam
c) The slightly unseasonal rainfall and inflows received in November and December 2017
The BLACK DASHED line is the projected storage assuming low inflows like last year.
The RED Dots are the actual weekly storage of the system.
“Dead storage” is the water in the dams that cannot be accessed due to pumping limitations. Berms and temporary pumps are being installed to reduce the dead storage.
The reason that Misverstand dam is excluded is because it is a relatively small dam with only about 10 million m3 storage (of which about 80% was unavailable due to pumping limitations) which serves the West Coast and some irrigation.
The totals include the following dams: Theewaterskloof, Voëlvlei, Berg River, Wemmershoek, Upper and Lower Steenbras.
For those who need a bit of info on stochastic modelling:
The probabilistic trajectories were derived using stochastic modelling. “Stochastic” means being or having a random variable. A stochastic model is a tool for estimating probability distributions of potential outcomes by allowing for random variation in one or more inputs over time. The random variation is usually based on fluctuations observed in historical data for a selected period using standard time-series techniques. Distributions of potential outcomes are derived from a large number of simulations (stochastic projections) which reflect the random variation in the input(s).
Its application initially started in physics, although it is now being applied in engineering, life sciences, social sciences, and finance.
The stochastic streamflow generator used in the system model was developed by well known University of KZN Professor Emeritus, Geoff Pegram.