Specialty Chemicals

The Planning Problem

Standard ERP finite schedulers were built for discrete assembly and relatively stable process routings. In specialty chemicals, that abstraction breaks as soon as changeovers are sequence-dependent: the duration, cost, and even feasibility of cleaning depend on the pair of campaigns you are transitioning between—not on a single average setup time attached to the next product. Clean-in-place (CIP) recipes, boil-outs, solvent flushes, and analytical release steps create a directed graph of transitions that generic tools rarely model. Planners compensate with spreadsheets and verbal rules ("never run this polymer after that acid," "this intermediate must clear the line within N hours"), which are correct until the next formulation change, alternate raw, or capital project rewires the plant.

Multi-stage formulations and co-product flows make the gap worse. Exploding a BOM is not the same as optimizing a route that branches on quality, reworks an off-spec intermediate, or diverts by-product to a recovery tank with its own compatibility rules. When optimization cannot see those structures, the schedule optimizes a cartoon of your plant. Tank farms add another layer: shared manifolds, heel policies, and storage classes mean a schedule that looks feasible in a Gantt chart fails on the floor because a heel is committed, a header is blocked, or an environmental hold delays release. Safety-stock targets in standard inventory modules cannot express "this tank cannot follow that campaign without an extra wash."

The predictable outcome is chronic expediting, last-minute campaign swaps, and conservative padding that hides capacity. WonForge encodes sequence-dependent changeovers, formulation depth, tank farm logic, and cleaning policies inside the optimization model so the plan respects the same physics and compliance rules your operators enforce—then searches for margin, service, and throughput under those constraints instead of pretending they do not exist.

Viscosity jumps, thermal profiles, and purity classes add sequence logic that is invisible to item-based ERP scheduling: the same SKU family can require different line preparation depending on the prior campaign's residual film, cooling state, or analytical category. Without modeling those transitions as first-class constraints, "optimization" devolves into freezing a narrow band of sequences that planners trust, which forfeits throughput and leaves money in solvent, energy, and lost yield.

Implementation Approach

Our chemical industry expertise begins with deep process understanding of complex formulation structures, multi-stage routing, and co-product relationships. We develop models that handle intricate formulations with multiple levels, reaction sequences, and material transformations. Implementation includes pulling data from your ERP system through our connector, integrating complex routing logic, and managing co-product and by-product relationships.