Tank Farm Optimization: Ghost Capacity Without CapEx

Introduction

Walk your site map and count the tanks. Now ask your planner how much of that volume they'd actually commit to in a plan. The second number is smaller, sometimes a lot smaller, and nobody is doing anything wrong.

That gap has a name worth using: ghost capacity. It's volume you own, paid for, and maintain, that your operation won't put into a plan because nobody can promise what will be coming into or out of that tank three days from now. It doesn't show up as a shortage. It shows up as shorter campaigns, extra washouts, and a plant that feels full while a third of the farm sits half empty.

Nameplate capacity and usable capacity are two different numbers

Every tank has a working volume on paper. Below that sits whatever headroom your process engineers require, and that part is not negotiable and shouldn't be: foam, thermal expansion, high-level alarms, spill margin. That's engineering, and it stays.

But on top of the engineering headroom, there's a second layer that nobody specified and nobody owns. It's the planner's margin, held because the plan they're working from can't tell them what's actually going to land in that tank mid-week. So they leave room. Not for a physical reason, but for an information reason. The first layer protects the process. The second one just protects the planner from a surprise the system was never able to warn them about.

Below the working volume sits engineering headroom, which is real and permanent. Above that sits planner's margin, held against uncertainty rather than physics. Those are different things, and only one of them has to be there.

The barrier is the tool, not the talent

Planners in this business are expert at Tank Tetris. They know which header is tied up, which tank drains slowly, which grade can't follow which. The problem isn't judgment, it's that they're being asked to hold a moving picture in their head while the system they've been given shows them a snapshot.

Most ERPs treat a tank as a storage location with a quantity, not as a level that rises and falls over time as production fills it and the downstream draws it down. They also don't carry the sequence-dependent rules that decide whether a transition is even allowed, so the planner can't ask the obvious question, which is "if I run this campaign, where does that leave tank 4 on Thursday?" When the tool can't answer that, there's exactly one defense available, and it's to leave room. Padding isn't a bad habit. It's the only rational response to a blind spot.

Planners pad because the system shows them a snapshot of a tank, not a level moving over time. When you can't see Thursday from Monday, holding room back is the correct move.

What changes when the plan can see the tank

Here's the part that actually recovers the volume. In WonForge, tanks aren't storage locations, they're constraints with levels. Each tank carries its capacity and its minimum, including whatever safety headroom you set, and the optimization tracks what fills it and what draws it down across the whole horizon. The plan it produces respects those limits in every period, and you can see the projected level of each tank over time rather than a single current number.

Two things follow. First, the plan can't propose something that overfills a tank or drains one it needed, because those limits are hard constraints, not warnings to catch later. Second, and this is where the ghost capacity comes back, the planner no longer has to pad against the unknown, because it isn't unknown anymore. The safety headroom you specified stays exactly where you put it. What goes away is the extra room held purely because nobody could see what was coming. You're not filling tanks fuller than your engineers allow. You're stopping the plan from pretending there's less room than there is.

Tanks are modeled as levels with capacity and a minimum you set, tracked across the horizon. Your safety headroom stays; what disappears is the padding held against a blind spot, because the blind spot is gone.

This is capacity you already paid for

The appealing thing about ghost capacity is where it comes from. It isn't a new tank, a permit, a construction schedule, or a capital request that has to survive a budget cycle. It's volume already sitting on your site, already maintained, already insured, that the plan simply wasn't willing to use.

And the recovery compounds, because tank room is what buys campaign length. More usable intermediate volume means the upstream can run longer before the downstream has to catch up, which means fewer transitions, fewer washouts, and less of the week spent cleaning instead of producing. The capacity itself is the smaller prize. The campaigns it lets you run are the bigger one.

Ghost capacity needs no capex, permit, or build. And because tank room is what lets a campaign run longer, recovering it buys fewer changeovers too, not just more storage.

Frequently Asked Questions

What is ghost capacity in a tank farm?

Ghost capacity is the gap between the volume your tank farm has on paper and the volume your planning is actually willing to commit to. Part of every tank is engineering headroom, which is required and stays. The rest is margin a planner holds back because the system can't show them what will be entering or leaving that tank later in the week. That second layer is the ghost: capacity you own and maintain but never plan with.

Does optimizing tank levels mean running tanks closer to full than is safe?

No. Your safety and minimum levels go into the model as constraints you set per tank, and the plan is built to respect them in every period. Nothing proposes filling past the headroom your process engineers require. What changes is the padding held on top of that for uncertainty, which exists only because the plan couldn't see what was coming. Make the level visible across the horizon and that extra margin isn't needed any more.

Our ERP already tracks tank inventory. Isn't that enough?

Tracking a quantity and planning a level are different things. An ERP can tell you what's in a tank now. What it generally can't do is project how that level moves across the horizon as production fills it and the downstream draws it down, under the sequence rules that decide which transitions are allowed. Without that, the planner can't test whether a campaign leaves a tank blocked on Thursday, so they hold room back instead. Modeling the tank as a level with limits is what closes that gap.

Conclusion

Your tank farm is almost certainly bigger than your plan thinks it is. Part of every tank is engineering headroom that should never move, and part of it is margin your planners hold because the system can't tell them what's coming. The second part is capacity you already own, already pay for, and never use, and getting it back costs nothing but a plan that can see each tank as a level moving over time instead of a number sitting still. If you want to see how much of your farm is ghost capacity, Check Your Fit is a short, no-commitment call to find out.

Tank Farm Optimization: How to Unlock "Ghost Capacity" Without CapEx