How accurate does your measurement device have to be? The Water Board gives those numbers in the Fact Sheet at http://www.swrcb.ca.gov/press_room/press_releases/2016/pr12016_measurement.pdf; see the bottom of this post for the excerpt on accuracy.
When talking about new weirs, orifices, flumes, mag-meters, and acoustic Doppler devices, plus or minus (+/-) 5% accuracy is expected of new, properly installed, regularly maintained, correctly operated devices. What does that mean? If your diversion rate is measured at 1.00 cubic feet per second (cfs), then you would expect the true value to be between 0.95 and 1.05 cfs. If your diversion rate is 5.00 cfs, then the true value would be between 4.75 and 5.25 cfs. The total accuracy is 10%, we just don’t know if measured values are really up to 5% less, or 5% more than calculated.
New devices might actually have better accuracy than +/- 5%. Engineers never count on that because a bunch of factors, known and unknown, can stealthily make the accuracy worse. Accuracy also depends on the measurer – some are better than others, some are better trained and experienced, and most take the job seriously but some do not.
Of course, accuracy gets worse as measurement devices age. Why does this happen? There are a number of reasons:
- Settling, so the device is not level front to back, or side to side, or both
- Cracking, so water leaks out, or the cracked wall is not straight (planar)
- Wear, spalling, chipping, and other roughening in the device floor and walls
- The ditch fills in downstream, causing submergence
- Old boards that warp and leak
- Installed staff gages wear, making them harder to read correctly
The USBR Water Measurement Manual has 14 chapters, and all of Chapter 3 discusses accuracy in great detail. That’s the “Bible” of water measurement so we would expect it to be, well, accurate in its discussion of accuracy.
It is not clear to me yet whether the Board’s accuracy numbers are +/- values, meaning the allowed accuracy is +/- 15% for diversions less than 100 acre-feet (AF) per year, and +/- 10% for diversions greater than 10 AF per year. If so, that seems reasonable because that allows for some aging of measurement devices. Otherwise, the Board would expect measurement devices to always be in new condition for diversions greater than 100 AF per year or storage greater than 200 AF per year. That would be pretty expensive!
That brings up the subject of money – accuracy requirements hit your pocketbook. First you have to either install or pay for a measurement device to be installed. Hopefully the device will last 20 to 30 years, but high flows, getting walked on by cattle, freezing and thawing, settling faster than expected, and other events can wear them out faster. The replacement cycle might be 10 years for some diversions, or even 5 if wear and tear is bad.
This post may be more than most people want to read on the subject of accuracy. Still, it’s a lot shorter than Chapter 3 of the Water Measurement Manual!
That’s all for now, have a great rest of the week.