通常，校准模型需要调整数据，但有时校准过程会显示异常情况会影响测量。模型通过计算管道中的流量和水头损失来预测系统性能和压力。为了进行真正的校准，模型必须与流量和水头损失测量值相匹配。一个匹配另一个还不够好。误差有时会相互补偿，使不准确的模型看起来像是被校准过的。本文介绍了四个成功校准的案例研究。

Usually calibrating a model involves adjusting data, but sometimes the process of calibration reveals that unusual conditions affected the measurements. Models predict system performance and pressures by calculating flow rates and head losses in pipes. For true calibration, the model must match both flow and head loss measurements. Matching one of the other is not good enough. Errors sometimes can compensate for each other and make an inaccurate model appear to be calibrated. Four case studies of successful calibration are presented.