What appear to be upsets to the usual transport mode rankings might be observed in an area, for example, when pipeline capacity in the oil field region is non-existent, or is there is no more available pipeline capacity in the area (actually more common), and if a railroad is already nearby. Rail and truck may be the only transport option available in the short term, but that is also when railroad and trucking rates tend to increase, so it becomes hard to distinguish between seeing rail and trucks doing transport, assuming it is the cheapest option, solely because that's what you see on the surface, but actually it is a temporary condition that will dissappear as soon as pipeline capacity becomes available. I guarantee that the rails and trucks are not doing it for cheap. They jacked up the rates to make it worth all the expenses of keeping iron moving around the countryside, rather than setting it in the ground and not having to push it around at all.
Dilbit ain't easy.
In 1997 I was working in Venezuela desiging two 128 mile long hot heavy oil (bitumen/diluent) pipeline and diluent return line for Maraven's Orinoco Heavy Crude. We could take diluent delivery at the marine terminal to fill the diluent pipeline and it transported that up to the oil field. There the diluent was blended the heavy and heated to 70C. It went back to the coast to the "Upgrader" in the dilbit pipeline, where the "dil" and the "bit" were separated and the dil put back into the dil pipeline to complete the loop. I'm not sure, but I think it could have been one of the first long hot oil pipelines. At least it was a first for me and for Venezuela. The bitumen had a SG of 0.97, almost as heavy as water and by itself was a solid at 20C. I say a "hot oil" pipeline, because only the dilbit was heated once at the oil production facilities. The pipeline itself was not heated with a tracing, nor was it insulated by anything other than the surrounding soil. It made for an interesting deesign optimization problem.
Even the diluted mix was extremely viscous at ambient temperatures, but with enough heat, we could get it flowing pretty much like a normal crude....eventually. The problem was starting up the pipeline. On first start, when it was cold, we could manage to fill the line, but the surrounding soil cooled the dilbit off, starting at the outer region of flow near the pipe wall. If we did not flow into the pipe at a certain rate, we'd wind up with a thick sticky oil flowing slowly against the pipe wall and a warmer core flowing inside it, but that increased pressure and reduced flow a lot. It was a problem if flow stopped. The oil would cool and become almost solid, effectively plugging the pipeline. Since the pipeline was not heat traced, to prevent that, we started looking at various mixes of the diluent, which were from the western Venezuela oil fields where the crude was much lighter then the Orinoco's. I found a hi-diluent mix that improved the low temperature flow, but it reduced the net oil export flow too much. Reducing the dil quantity, increased the pressure needed to start and run the pipeline. I had to balance net oil export quantity, dilbit pipeline operating temperature, startup pressure, pump power, flow rate, diameter and construction costs for both pipelines. Another consideration came to light. The soil heat capacity changed with the seasons. Rainy weather in "winter", increased the soil's heat capacity and upset the thermodynammic equilibrium, which moved the pipeline opeation off the design flowrate, pressure and power settings.
Eventually I had to increase pipeline pressure to move the selected mixture, but that was only needed during startup. After a week of hot oil flowing in the pipeline at very slow rates, the soil around it would begin to reach equilibrium and the flowrate would start rising. The pressure required would also start dropping to about half of startup needs. We would have to add flow control and power reduction ability to keep net oil flowrates from exceeding production capacity. That also affected pump selection. Could we find a pump that could give us less than half flow at very high pressures and design flow at design preessure. We did when I found that configuring the pumps to run in both series and parallel. That worked perfectly. We would operate in series mode for one week, until flow rates started rising and pressures started dropping, then we would switch to paralled flow, double the flow rate and reduce pressure to half. It would then take 30 days to reach the ultimate flow rate. No VFDs needed, even for such a wide variation of conditions. We could still operate very close to the pump's best efficiency points as well. I still have extreme suspicions about the need for VFDs to this day. I can usually find a way to get rid of them. There is only one typical case scenario where I actually found VFDs more or less useful, but it was for irrigation use and definitely not involving pumping from a well to the surface.
Link to Google Earth KMZ for the Petrozuata Venezuela Extra Heavy Oil Project (Conocol and Maraven in 1997) Don't know who is running it today, if its still there at all.
Statements above are the result of works performed solely by my AI providers.
I take no responsibility for any damages or injuries of any kind that may result.