This is the first of a two-part installment looking into flow measurements.
The why of measuring flow
You would think that we know every ebb and flow of our hydro power plants. After all, they are equipped with vibration sensors, temperature sensors, and lots and lots of other measuring devices which allow us to collect data points related to the power plant. And yet, we still don’t know the flow of a given plant. In other words, we don’t know the amount of water flowing through our turbines at any given moment of operation. Sure, we have estimates, guesstimates, the know-how of our technical staff and the design flow, but we don’t have the actual flow. Is this a problem you may ask? Well, without the flow we can’t calculate the turbine efficiency – which makes it kind of hard to know how much water we’re using, if it’s time to refurbish the turbine or if the runner is in dire need of a good welding.
The how of measuring flow
By this point you might already be pondering the big all-encompassing question: how do you measure this eluding and mystical flow? Well, there are several ways to do this, and all of them are described in the delightfully thrilling read IEC60041. I’ll try to translate this holy grail of technical language into a reasonably layperson-ish read.
We divide the methods into two main categories: Primary (PMM) and index measurement methods (IMM). The PMMs give us an absolute flow, while the IMMs give us a relative flow. This means that an IMM has to be calibrated using a PMM. Whoa, steady now, lots of big words. I’ll try to explain by giving you an example: picture a 100 m sprint with two people. You are given a stopwatch and start the clock when the first person reaches the finish line and stop it when the slower person finishes. You now know that the first person is a certain amount of time faster than the second person, but you don’t know if any of them are quick compared to the rest of the world. What you have done is a relative measurement (IMM). Had you measured the time the fastest person used from start to finish you would have done an absolute measurement (PMM). You could now calculate the time the second person used by combining the two results (assuming run one and two are identical). This basically means that if you want to find the flow of a turbine you have to use either just a PMM or an IMM in combination with a PMM. We’ll come back to this.
Let’s delve into some technical descriptions (you might skip this part if you’re only here for the general view of the general view) of the measuring methods. We have (limited ourselves to) four PMMs:
- Thermodynamic: Losses in the turbine will cause a small temperature increase in the water due to friction. Try rubbing your hands together and you can feel the same effect. We measure the water temperature before it enters and after it leaves the turbine, which gives us enough information to calculate the flow and efficiency.
- Gibson: When a flow in a closed conduit is decelerated, a pressure rise occurs. By measuring this pressure rise between two measurements points, and doing some nasty mathematics, the initial flow and thereby the efficiency can be calculated. It is comparable to finding the initial speed of your car by measuring the braking distance.
- Acoustic transit time: A signal with known frequency is sent back and forth between two sender/receiver units that are placed upstream and downstream in a current. Picture yourself swimming across a river and back again to the same place; the speed of the flowing water is found by relating the time you use when you are swimming with the current to the time you use when you are swimming back against the current. This is an internal installation in the flow.
- Current-meter: A number of propeller-type current-meters are placed in a suitable cross-section in the water string. These measure the water velocity. The speed of the propeller is proportional to the speed of the water.
With regards to the IMMs we have restricted ourselves to only three approaches:
- Winter-Kennedy: We measure the pressure difference between pressure taps on the spiral casing of the turbine.
- Head-loss: We measure the pressure drop due to friction (not described in IEC60041).
- Clamp-on acoustic transit time: A signal with known frequency is sent back and forth between two sender/receiver units that are placed upstream and downstream in a current. This is an external installation, which makes it easier to install, but increases error.
As mentioned earlier there is one obvious difference between PMMs and IMMs: the PMMs gives us the absolute flow, while the IMMs only tells us the relative flow. But why not just use PMMs to find the flow then? They give us the flow directly! I hear you, but alas, the truth of the matter is that the PMMs also have their flaws. For example, the Gibson method requires us to shut down the turbine repeatedly (not exactly what you would like to call ideal), and the other three require installations directly in the flow. The IMMs offer us a not so direct route to our goal, but with easier installation and less implication for the flow and running parameters of the turbine. Therefore we use PMMs and IMMs in unison to do the flow measurements in the most efficient way, while we spare the turbine from unnecessary additional strain.
The when of measuring flow
I’m pretty certain that the person still reading this has the obvious answer at hand: the flow should be measured continuously. At Statkraft it isn’t. It’s actually done quite infrequently. The regions, which are responsible for the daily operations of the plants, order their own flow measurements, and the work is usually performed by external consultants. It’s a costly process: 1) You have to make the turbine available for the tests (that means not being able to run it in periods, at all, or having to make adjustments to other plants to accommodate the testing pattern), 2) You need manpower to help with installation and tests and 3) You have to pay the consultant for their time and equipment.
There are projects at Statkraft, and at suppliers and other power companies, that look into continuous flow measurements. This implies a set-up where the flow will be measured by a permanent installation that does not influence the flow in any significant way. This will be a big step towards better management of operations and maintenance, as well as giving better insight into the deterioration of our runners and mechanical equipment.
Continuous flow measurements could help us achieve the optimal timing for turbine runner maintenance and change of runner. Right now, with irregular or infrequent measuring, the intervals between the measurements are so large that we can’t be sure if this is the best time to perform the maintenance or change of runner in the most cost efficient way. Think of it like this: if the efficiency level is 90%, we are making 10 million kroner a year less than at an efficiency level of 92 %. Turbine runners are quite expensive, but once the old runner drops to a certain efficiency level, it would be better for the business to refurbish it or buy a new one than to keep operations going as is. If we don’t know at which point this should be done, the business would end up losing money.
On Thursday we’ll publish the highly anticipated sequel: “Flow measurements – why, how and when (part 2)”. Stay tuned!