Answering the Question: Is a Surge Analysis Required?
by David L. McPherson, P.E., M.ASCE, (Ph.D. Candidate, Alan Plummer Associates, Inc., 1349 Empire Central, Suite 1000, Dallas, TX 75247. Email: dmcpherson@apaienv.com) and Christopher Haeckler, A.M.ASCE, (Alan Plummer Associates, Inc., 1349 Empire Central, Suite 1000, Dallas, TX 75247) Section: Pipeline Planning and Design, pp. 929936, (doi: http://dx.doi.org/10.1061/41187(420)85)
Access full text Purchase Subscription Permissions for Reuse
Document type: 
Conference Proceeding Paper 
Part of: 
Pipelines 2011: A Sound Conduit for Sharing Solutions 
Abstract: 
In practice, the hydraulic transient analyst will inevitable be posed with the following question, "Is a surge analysis required?" The discussion herein proposes an easy way to make this decision. In hydraulic modeling, the St. Venant Equation is used to characterize the relationships between hydraulic head, flow rate time, and distance along the pipeline. A term within the St. Venant Equation holds the answer to "the question." This acceleration term is characterized by the momentum equation. The use of acceleration will help the analyst understand the sensitivity of the model to hydraulic transients which will potentially save time and effort as well as help the analyst avoid the use of an overly complex model for an analysis. The acceleration term considers the interrelationship(s) between the velocity change, the time change, as well as the length of the pipeline system. This paper will show where this acceleration term is within the St. Venant Equation and explain how it can be coupled with various assumptions to answer "the question". Along with the acceleration head parameter, the transient analyst can utilize a matrix along decision criteria, which are developed within the paper, to assess if the transient behavior of the system warrants consideration for full dynamic "surge" modeling. This matrix will also help with the decision of what type of hydraulic model to use. Is a steady state model adequate? Is a quasidynamic or fully dynamic model required? The use of the acceleration term will be introduced as a decision tool for this line of questions as well. The concept (model simply for results and/or model to fit) will be fleshed out within the paper and a discussion with proposed decision criteria for both questions: "Is a surge analysis required?" as well as "What type of model should I be using?" will be presented. 
