The Whizzy Stick

The canals are around 200 years old, but the technology we use to manage the water in them is ‘state of the art'.

Much of the waterways are monitored by computer technology known as SCADA (because Supervisory Control And Data Acquisition is a bit of a mouthful!). We use the SCADA to measure water level. When we know what the water levels are, and when we know how much water is moving around and where, we can calculate water budgets, plan for shortages or react to flooding. We can also use it to remotely operate the many automatic sluices, adjustable weirs and pumping stations around the country. It can send out alerts to telephones or pagers when something goes wrong. In this way, we maintain the water levels in the canals for safe navigation.

But to do all of this well, the equipment must be accurate. Part of my job is to make sure it is accurate. I have been trialling some Global Positioning Satellite (GPS) surveying equipment to get that accuracy. I call it the Whizzy Stick because nobody that I can find has told me a concise, meaningful name for it – perhaps you know better?

This is how the SCADA works

A pressure transducer, or level sensor, is fixed to a point under the water. As the water level rises or falls, the water pressure on the sensor changes. That pressure change is converted to a change in water level height in millimetres. Every canal pound has a nominal water level, and it is most often the height of a weir crest. So, we want to maintain the water level at close to the height of the weir crest. If I know the height of the weir crest, and I know the relative height of a fixed point next to our level sensor, and if I then measure the distance from that fixed point to the surface of the water, I can calculate where the water level is at the weir; either so many millimetres going over it to waste, or so many millimetres below and we want more water. The SCADA tells us the height of the water in relation to the weir crest. You can see that for the SCADA to be accurate I must know with accuracy the difference in height between the weir crest and my fixed point at the level sensor.

When the weir is close by this is easy; I can just use old fashioned surveying kit that has not changed much since Roman times – and we know how good they were at building flat, straight roads. But when the weir is a long way off it is much harder to do. The Whizzy Stick may be the answer.

We've a team based at Leeds who have, for a long time, been using GPS surveying to map our weirs, reservoirs and so on. The device takes readings from multiple satellites and produces a height measurement of millimetre accuracy. It occurred to me that it could provide me with the accuracy I need to calibrate the many SCADA sites we have where the controlling weir is a long way from the level sensor. To that end I have surveyed all of our SCADA sites on the Kennet & Avon Canal and the area around Braunston Summit on the Grand Union Canal (where automatic, SCADA operated pumping stations are so important), and also some of the SCADA sites on the Mon & Brec and Swansea Canals. There have been mixed results as we shall see.

The GPS surveying device, when it can locate enough satellites, and get a good mobile phone signal is able to tell me the height of a point to within 15 to 20 millimetres. When you consider the distances involved this is quite remarkable; and for my purposes, for the most part, good enough. The problems come when that level of accuracy is not achieved. Imagine that the device returns a height measurement that is out by 40 millimetres, and so after calibration the SCADA tells a pumping station to top up a 3km pound by 40 millimetres when in fact it is at weir crest height. That would result in a lot of water pumped to waste, and a big electricity bill to boot. There are ways around this, but the potential for inaccuracy needs to be understood.

I have also discovered that in more remote places, like the Mon & Brec Canal in the Brecon Hills, the device cannot get a reading because it needs a good mobile telephone signal. It also cannot get a reading if there are overhead trees or hills nearby that prevent a clear line of sight to the overhead satellites.

An operative checking a permanent SCADA level sensor on a canal pound.

Overall the trials have been a success

For instance, it only needs one person to operate, whereas the old-fashioned surveying always requires a minimum two people. Where appropriate the device has given us excellent accuracy over long distances. We now know that we need to apply careful checks and offsets to achieve the accuracy we want, and not just rely on the number the device tells us. We have also been able to add to the mapping knowledge that has been accumulated by our dedicated mapping team.

We know that for our more remote canals we are better off relying on traditional methods of height measuring, even though it may require more manpower or achieve less accuracy. With the knowledge gained, the Water Management Team's Water Engineers will be using the device as part of their SCADA site inspections program, building up a database of height measurements for the future. Come and talk to me if you see me out with the strange looking Whizzy Stick – I will be glad to answer any questions and tell you about all the wonderful things we can do with SCADA.

By Mike Wheeler, Water Engineer for the South