Dear sirs

Most of the hydroelectric power plants i'm working on use speed pickups to measure the rotational speed of the generator. Nevertheless, there are a couple of units that use frequency transducers (using voltage reference from the generator bar) as the main speed reference, and using speed pickups as secondary signals, used only in the machine start up and in case of failure.

From what i could see about this signals, frequency transducers deliver a much more clean and precise speed measurement, compared to speed pickups which deliver a noisy and variable measurement (around +/- 1% of actual speed).

My question is, based on your experience and knowledge, would there be a significant change in the control of our machines if all the speed signals were measured by frequency transducers? Or is this change irrelevant for the speed controller?

Thanks for your time. I'd be glad to provide any necessary additional information in request.

 

delirios,

How are you measuring the accuracy of the speed pick-ups--by a value on a display or a meter? Are you looking at the output of the speed pick-up directly using a DMM (digital multi-meter with frequency-measurement capability), or the scaled feedback on a display?

How do you measure the accuracy of the frequency transducer--by looking at the output of the transducer using a DMM, or the scaled feedback on a display?

Are the control systems all the same model and vintage, or are they different?

What is the output of the frequency transducer: frequency (Hz), or voltage, or 4-20 mA?

I think you'll find most control & protection schemes will use one or more speed pick-ups as they are generally much more reliable. Most voltage-measuring devices (PTs--Potential Transformers) have a fuse or fuses on the high side. And, as you've noted, the control systems that do use the frequency transducers as speed signals DO have a speed pick-up as a backup.

I'm still curious, though, how you're measuring the accuracy of the two types of devices. Are you sure the systems with the high error are not the result of poor governor control/regulation when not synchronized? Do you see the same error on the same unit/control system when synchronized as when accelerating or when trying to synchronizing?

 

CSA

I'm measuring the accuracy of both speed pick-ups and frequency transducers by looking at a scaled feedback on a display.

We use the Sulzer DTL 595 governor on almost all of our control systems. We have the option of connecting to the governor and monitoring each signal through a serial port. When I start graphing speed signals (n1 for frequency transducer, n2 for speed pickup) I notice that n1 shows much more stable values, whether n2 has constant up and down peaks on each sample (small peaks of around 1-2% of actual speed).

This speed pickup oscillations are seen all the time, whether the unit is on start-up, synchronizing or synchronized to the grid.

My concern is that this small up and down peaks shown on the n2 signal might be making the governor to constantly send small opening and closing signals to the injectors, making more strain and wear in all the components.

 

Assuming that there are no problems with the sensors or installation (correctly screened cable), it is much safer to use the pickup over the frequency transducer. One thing you should consider is that if there is a loss of excitation, then what will the governor do when it senses no speed input from the frequency transducer? Will it open the valves to drive up the speed?