You may not know that South Africa is a water-stressed country, or that it requires a scientific mindset and innovative solutions to keep our taps flowing. Meet a band of unlikely rock stars…
In their quest to manage South Africa’s limited fresh water supply, a small band of hydrometric technicians has helped develop the most sophisticated water-flow measuring tool ever built. Their contribution to the design of Sontek’s revolutionary Acoustic Doppler Profiler (ADP) is deemed so valuable, in fact, that they’re regarded by overseas counterparts as the “rock stars of the hydrological world”.
Mildly hyperbolic? Not at all. Consider this: whereas gold and platinum may be extremely valuable commodities, fresh water is without doubt our most precious natural resource. After all, if our rivers and dams were to run dry, we can say goodbye to urban development, food production and our ability to generate electricity. The consequences to our very existence are unimaginable.
In a water-stressed country such as ours, it’s critically important that we take care of what little we have, and manage it well. This heavy responsibility falls squarely on to the shoulders of the Department of Water Affairs (DWA), which in turn relies on data collated by its team of hydrometry specialists.
Frans le Roux, the department’s area manager for hydrometry in Gauteng region, occupies an office on the banks of the scenic Boskop Dam just outside Potchefstroom. The foyer’s left wall is dominated by a massive colour photograph of the oldest flow gauging weir in the country. Built in 1904 on the Mooi River, upstream of Potchefstroom, the valuable data it provides is still being used to this day.
But it’s the larger-than-life graph taking up the opposite wall, titled “100 Years of Flow Monitoring, Gauteng Station”, that really catches the eye. Prepared by Le Roux to mark the centenary of the Mooi River weir, it faithfully depicts flow rates measured by the weir from its inception right up to 2004. Intent on adding a human perspective, he’s included photographs of historic events that correspond with the dates of significant recorded events. It works rather well, bringing the area’s hydrological history to life in a way that no dry array of data could hope to match.
Picking out the major flood peaks is a fascinating experience: 1904 – Einstein proposes his theory of relativity; 1909 – Robert Peary becomes the first man to reach the North Pole; 1914 – start of the First World War; 1944 – war continues to wreak havoc around the world; 1976 – the Soweto riots begin; and finally, 2000 – the year the cricket match-fixing scandal broke.
Accuracy is everything
South Africa has – needs – one of the most advanced hydrological monitoring departments in the world. Says Le Roux: “We have a proud history and lots of locally developed skills attuned to our specific demands. Because of this, we regularly host our international counterparts, who want to see what we’re doing.”
In essence, the DWA technicians live for accurate information. “Short-term data is valuable, but long-term monitoring is critical,” says Le Roux. “Before you can make good assumptions, such as where to locate farms or new settlements, you need at least 30 years’ worth of flow data. A big swearword for us is ‘data gap’, because that means we have to start guessing.”
As Le Roux sees it, South Africa’s rivers (the country’s primary water supply) may be described in just three ways: “Too little, too much or too dirty.” But this is the least of his worries. What’s of more concern is that only 20 per cent of the land surface accounts for about 60 per cent of our annual rainfall. Oh, and its distribution is also highly seasonal.
Consider Gauteng’s geographical location. The economic hub of the nation, it’s home to about one-third of the population. Johannesburg is the only metropolis in the world not located on a coastline or major river – and let’s not forget the mining industry, the large coal deposits and the numerous, water-guzzling power stations that keep our lights on.
“From a water supply standpoint, Johannesburg is located in the wrong place,” says Le Roux. “The Upper Vaal catchment area simply can’t provide enough, so we have to transfer water from all over the place to supply demand there. Making this happen requires a huge, highly complex infrastructure. This is where water runs uphill.”
Then there’s flooding. Although the primary function of our dams is to store water, they’re also a great help when it comes to flood management. “If a flood is going to happen, we can make it ‘flatter’ by pre-releasing water from some dams. We can’t keep a flood back, but we can smooth out the flow to help minimise damage,” Le Roux says. “When the flood has gone, we ideally want to end up with a full dam. Without accurate information on flow rates in the system, this becomes almost impossible.”
As the custodians of the country’s fresh water, Le Roux and his colleagues are tasked with planning for our future needs, which means knowing where to build new dams, including their optimum size. “You need a dam large enough to supply demand, but small enough to keep costs down. It’s a trade-off,” he explains.
The Wolwedans Dam near Groot Brak in the Southern Cape is a good example. Built in 1990 at a cost of R32,8 million, it was designed according to historical hydrological records. “If the flow rate was underestimated by 10 per cent, it would mean there was more water available than envisaged, and the dam would be bigger than required. This would represent a waste of 36 per cent, or R11,8 million in monetary terms. Apply the same error to the new Berg River dam and you’re looking at an overspend of R512 million. That’s why I’m so pedantic; I understand the ramifications.”
Out with the old
Sontek’s fancy new tool has undoubtedly revolutionised the hydrological profession. Le Roux reckons its impact is comparable to the effect of mobile phones on the communications industry. “The technology behind telephones remained virtually unchanged from the time of Bell’s concept right up to the advent of cellular communications. The materials and technologies may have improved but the basic working principle remained the same. Well, the same thing has happened here.”
Le Roux explains the basic principle of flow rate: “You take a 200-litre drum, turn on a hose and measure the time it takes to fill up. You then divide the 200 litres by, say, 120 seconds to get your flow rate. This is called the ‘volumetric method’. The thing is, how do you conduct a volumetric measurement on the Orange River in flood?”
Although measuring the normal flow of our river and canal systems is logistically challenging, it has never posed a serious problem, thanks in part to South Africa’s extensive network of purpose-built gauging weirs. However, they can’t cope with the variables introduced by flooding. Until recently, rapid flow rate fluctuations could be monitored only using manually operated current meters.
This meant actually wading into the water with an instrument that, although refined over the years, worked fundamentally the same as the first propeller-type current flow rate meter invented in 1790. “The only major technological change happened in the 1920s, when the counter was upgraded from mechanical to electrical.”
Water velocity over a cross-section of river is never uniform. It flows strongest in the centre at the surface and progressively slows down on the sides and bottom as a result of friction. Mechanical current meters can measure velocities only at a specific point, so multiple measurements have to be taken to create an accurate velocity profile. This involves separating the cross-section into 25 to 30 segments, lowering the instrument to various depths multiple times at each point, then crunching the numbers. It’s a tedious, back-breaking exercise.
Recalls Le Roux: “In the 1980s, I measured the Vaal in flood by dangling the instrument beneath a cableway. Aside from taking at least six hours, it was also physically dangerous. Lowering the heavy instrument with its large stabilising weight into the water while whole trees were hurtling past was no joke.”
In with the new
Understandably, when the first-generation Acoustic Doppler Profiler (ADP) designed specifically for rivers – they were originally developed for oceanographic applications – became available in 2003, the gurus at DWA sat up and took notice. Le Roux remembers “We used one for the first time in 2005. It featured only three acoustic transducers, and although I wouldn’t describe it as crude, we were pretty anal about accurate measurements, and figured it could be made to work much better.”
Fortunately, South Africa’s gauging weir system is one of the most extensive in the world. This put Le Roux and his team in the ideal position to test the ADP and supply Sontek, the instrument’s American manufacturer, with quality test data that other countries would battle to match. They promptly formed a close working relationship.
In 2008, work began on the second-generation ADP, and the DWA was involved right from the start. One major improvement was the inclusion of two sets of four-cluster transducers; one low-frequency (1 MHz) set for penetrating deep water and the second set, emitting higher frequencies (3 MHz), earmarked for the provision of good resolution in shallow areas.
Another useful modification was the inclusion of a central 500 kHz echo sounder for generating accurate depth readings. Computing power was also dramatically improved; rather than having to be controlled by a computer, the new version became a computer.
One big hurdle remained: being based on a differential GPS system, it would work only on wide rivers where accuracy wasn’t an issue. Says Le Roux: “On a river narrower than 75 metres it became too inaccurate – and this effectively eliminated all of our rivers.”
The obvious solution – for the DWA team, anyway – was to fit a Real Time Kinematic (RTK) GPS. Using two receivers, a fixed base station and a rover unit, it offered an accuracy of 1 cm.
However, Sontek were adamant it couldn’t be done. “We didn’t agree at all,” recalls Le Roux. “We bought some beers and asked our surveyors to bring along their RTK system and some wire. Using a multimeter to see which wires could be cut, we removed the differential GPS from the unit and connected the RTK. After a bit of fiddling, it worked perfectly. We then sent them a picture showing the wire connections and all our data. They implemented the modification immediately.”
All in all, the development process took two years and apparently cost Sontek a whopping $80 million (over R640 million at current exchange rates.) But Peter Wigzell, Sontek’s local representative, considers it money well spent. “Without these guys, we would never have been able to get our technology to the current level,” he says.
Nowadays, the hydrology team feels pretty good about its work and especially about its role in refining the ADP technology on which they rely so heavily. As Le Roux says: “It does everything automatically and can correct any aberration you put in its way. All we need do is put it in the water, pull it across the river on a line and press stop at the other side. Now, for the first time, we can measure stuff like never before. It’s simplified our lives enormously.”