During model development and calibration, errors in field measurements propagate to uncertainties in parameter uncertainty. Water distribution water quality models are typically calibrated using tracer test data. Here, we evaluate the uncertainty resulting from the test assuming demands are known exactly and only the wall coefficients are to be estimated. A meta-heunstic search engine, Shuffled Frog Leaping Algorithm (SFLA) is linked with hydraulic and water quality simulation software EPANET to estimate the pipe wall decay coefficients. A first-order approximation is applied then to estimate the parameter uncertainty. This methodology is applied on a mid sized example network. Results show that with tracer injection parameter can be accurately estimated. It is also found that parameter estimation can be more accurate if the magnitude of the wall decay coefficients is higher.