Production Engineers are asked to provide estimated productions and operational guidance for dredging plant that works in a rapidly changing and dynamic environment. These engineers use models to apply scientific principles and engineering practices. The models must have three properties; they must be tractable, discernable and informative. Tractability means the ability to quickly setup and run the model, as the engineer has to quickly evaluate multiple scenarios. Discernable models use parameters and inputs likely to be known to the engineers and eschew parameters that are unknowable or unimportant. By informative, we mean that the output of the model must provide a direct answer to the specific engineering question, provide other outputs that can be used as operational guidance, and to check and calibrate the model. To satisfy these requirements the models are generally analytical, based on fundamental or empirical equations that reflect the underlying physics. The paper contrasts such engineering models to those used in design engineering and to scientific models used for research purposes.