Transient pipeline optimization is in much the same immature state as pipeline simulation was 30 years ago. Both started with new technologies, and glimmers of ideas about how the technical advances could actually be used in practice. Ideas currently abound on ways to use transient optimization. Just as in the early days of pipeline simulation, it seems certain that some of our current ideas will pass the test of time, while other ideas may prove less useful. Conversely, it is also certain that we will find novel ways beyond what is already conceived. In this talk we hope to educate the audience on some current projected uses of transient optimization, and at the same time encourage discussion among potential users of the technology on what they need and what they do not, and why.
When transient pipeline simulation began to become available a few decades ago, two questions immediately arose. First, given the fledgling state of the art, what pipeline features and operations could we actually, technically, simulate? Second, were any of the things we could simulate actually useful? At the start, only a limited amount of thought had been given to either of these matters, because they were brand new. But with time, practitioners advanced the technical art of simulation, and pipeline companies thought of a widening variety of practical uses for these technical capabilities. These two processes were complementary; as operators thought of new practical uses they spurred new development within the simulation community. And as the simulation community enhanced the scope of simulation, these enhancements encouraged pipeline operators to be even more creative in using said enhancements. Transient optimization is in a similar position. A variety of remarkable technical capabilities have been demonstrated, many aimed at helping with known issues that arise at real pipelines. (E.g. -) However every pipeline is different and the actual range of potential uses is certainly broader than the examples so far published. Continued industry-wide feedback is needed to guide technical development. Conversely, many potential industry users of this technology would benefit from thinking outside the box and applying the technologies in ways not currently envisioned. This paper was originally envisioned as a workshop wherein potential users and developers could have spirited discussion regarding what is needed and what is possible. Although the material is instead being presented as a paper/presentation, our goal remains to provide a base for general discussion rather than to promote particular capabilities or methodologies. To provide this basis, in this paper we first explore some of the mathematical problems that Transient Optimization can solve. We show some ways this might be used, and how this differs from both simple transient simulation and from steady state optimization. We will skim through uses suggested in the literature. To give our discussions some grounding in reality, we go into much more depth on one real example from OMV.