Capacity planning when the pipeline is the constraint
Search for capacity planning software, and you will mostly find tools for staff rotas, server load, and project resourcing. That is what the term means to most industries. Add a person, add a server, add a shift. Capacity flexes.
A pipeline does not work like that. Its capacity is fixed by steel in the ground. You cannot add a lane at short notice. So when nominations exceed what the pipe can physically carry, someone has to decide whose volume moves today and whose does not. That decision, made well or made badly, is what capacity planning actually means in this industry.
What capacity planning covers on a pipeline
Three numbers sit at the heart of it, and they rarely match.
- Contracted capacity. What each shipper is entitled to move, under their agreement.
- Nominated volume. What each shipper has actually asked to move, for a given period.
- Available capacity. What the pipeline can physically carry right now, given linefill, maintenance, and everything else already committed.
Most of the time, these three numbers behave. Nominations sit inside contracted capacity, and contracted capacity sits inside what the pipe can carry. The planning is routine.
The interesting part, and the hard part, is what happens when they stop behaving.
The scramble
Something changes. A plant goes down, and a shipper needs to move volume they had not planned for. A cargo gets swapped at short notice. Weather closes a route, and everything behind it has to find a new one. Nominations across the network suddenly exceed what is physically available.
Someone now has to apportion the shortfall. Fairly, defensibly, and fast, because the pipe does not wait while people find the right spreadsheet.
This is where a lot of pipeline operations quietly fall over. The contracted capacity lives in one system. The nominations arrive by email or portal. Someone works out the available capacity by hand, from linefill data that is already a day old. Apportioning the shortfall means pulling all three together under pressure. Then someone has to explain the outcome to shippers who each think their volume should have taken priority.
NWSOLC faced exactly this. Long-term LNG cargo forecasts, pipeline nominations, spot cargoes, cargo swaps, and weather disruptions all had to be managed for multiple operators accessing the same system, on top of an existing platform that was slow and short on transparency. The scramble was not an occasional bad day. It was the normal operating condition.
What better data changes
None of this requires predicting the future. It requires the three numbers to live somewhere they can actually be compared, in time to act on them.
- One place holds all three numbers. Contracted capacity, nominations, and available capacity sit on the same platform, not across a contracts database, an inbox, and someone’s memory.
- Apportionment is a calculation, not an argument. When nominations exceed capacity, the rules for who gets cut and by how much are defined in advance and applied consistently, not negotiated fresh under pressure every time.
- Every shipper sees the same numbers. Self-service access to nominations and allocations means a dispute starts from agreement on the facts, rather than a debate about whose spreadsheet is right.
- A change is a same-day job, not a days-long one. When the inputs update, the picture updates with them, so the team spends its time on the decision, not on rebuilding the numbers that inform it.
That last point is where NWSOLC’s result is worth naming directly: the same automation that removed the scramble took their scheduling process from days to hours.
Complexity does not disappear just because the data behaves. Atlantic LNG runs nominations and allocation across four LNG trains, each with its own allocation rules, fed by gas from multiple fields across three pipelines, with losses and fuel gas accounted for on every stream, every day. None of that complexity went away when they moved it onto one platform. What changed was that a system upgrade delivered an order-of-magnitude improvement in processing speed for some of those calculations. The complexity became manageable instead of exhausting.
This is a planning problem before it is an allocation problem
It is worth being precise about where this sits. Once volume has actually moved, working out whose barrel or unit of gas went where is allocation, and we have covered that ground in our piece on pipeline management software. Capacity planning happens earlier, before the volume moves, when the question is still what the network can carry and who gets to use it. Get that decision wrong, and no allocation calculation afterwards will make it feel fair. If the aftermath is where you keep landing, whether numbers stop reconciling once the dust settles, that is worth a look at why variance analysis takes as long as it does, because the same disconnected data usually sits behind both problems.
Bring your worst nomination day
If there is a day on your network that everyone still talks about, the one where three shippers all needed priority at once, that is the best test of any capacity planning approach. Book a call, and we will match you with a partner who knows pipeline operations and can show you how your team would handle that day differently.
