19.09.2015 Views

Modeling

R184_CHI_Rules

R184_CHI_Rules

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Uncertainty Analysis 161<br />

area, New Jersey. (Water Science and Technology Vol 36 No 5 pp 141-<br />

148.(Anonymous, 1997).<br />

The American Water Works Association (AWWA) Engineering Computer<br />

Applications Committee indicate that true model calibration is achieved by<br />

adjusting whatever parameter values need adjusting until a reasonable<br />

agreement is achieved between model-predicted behavior and actual field<br />

behavior (AWWA Engineering Computer Applications Committee 1999). Once<br />

a model is considered to be calibrated, it can then be used, among other<br />

purposes, to estimate hydraulic characteristics of the real-world system at<br />

locations where measured data are unavailable or unknown, spatially and<br />

temporally.<br />

In the US, definitive standards to assess the accuracy of water distribution<br />

model calibration have yet to be agreed upon or established. However, the<br />

following calibration criteria have been suggested:<br />

1. An average pressure difference of ±2.2 psi with a maximum difference<br />

of ±7.3 psi for a good data set, and an average pressure difference of<br />

±4.3 psi with a maximum difference of ±14.2 psi for a poor data set<br />

(Walski 1983); and<br />

2. The difference between measured and simulated values should be ±5 psi<br />

to ±10 psi (Cesario and Davis 1984).<br />

According to the AWWA Engineering Computer Applications Committee<br />

(1999), ten sources of possible error could cause poor agreement between water<br />

distribution model values and measured field values. These sources of error,<br />

which provide a potential list of factors that can be adjusted during the modelcalibration<br />

process, are:<br />

1. errors in input data (measured and typographic),<br />

2. unknown pipe roughness values (i.e., Hazen-Williams C-factors),<br />

3. effects of system demands (distributing consumption along a pipe to a<br />

single node),<br />

4. errors in data derived from network maps,<br />

5. node elevation errors,<br />

6. errors introduced by time variance of parameter values such as storage<br />

tank water levels and pressures,<br />

7. errors introduced by a skeletal representation of the network as opposed<br />

to modeling all small-diameter pipes,<br />

8. errors introduced by geometric anomalies or partially closed valves,<br />

9. outdated or unknown pump-characteristic curves, and<br />

10. poorly calibrated measuring equipment including data loggers, tank<br />

water-level monitors, and SCADA systems.<br />

Model calibration entails adjusting model parameter values until an acceptable<br />

match is achieved between measured data and model-simulated values (e.g.<br />

pressures at the test hydrants, water levels in the storage tanks, flows from<br />

booster pumps, and pumpage from groundwater wells). The ten sources of<br />

possible error that could lead to model simulated values not agreeing with

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!