Skip to content

Refinery & Network LP Models

Building Planning Models That Reflect Your Refinery

Linear Programming is one of the principal mathematical optimization techniques used in refinery planning.

A well-designed LP model represents the physical configuration, operating constraints and economic relationships of a refinery and allows alternative operating strategies to be evaluated systematically.

DSCS provides LP model development, upgrade, validation and optimization services for refineries and interconnected downstream networks.


New LP Model Development

We develop refinery LP models based on the actual configuration and operating practices of the client’s facility.

Our work can include:

  • Data collection and validation

  • Refinery configuration development

  • Process-unit representation

  • Material balances

  • Product blending

  • Product specifications

  • Utility systems

  • Operating constraints

  • Feedstock and product economics

  • Model testing and validation

The objective is to create a model that provides a reliable representation of the refinery for planning purposes.


Existing LP Model Upgrade

Existing planning models often require modification as refinery configurations, operating practices, feedstocks and product requirements change.

We review and upgrade existing models to improve:

  • Model accuracy

  • Representation of refinery operations

  • Model usability

  • Economic relationships

  • Operating constraints

  • Solution quality

  • Planning flexibility


LP Model Validation

A planning model is only useful if it represents the refinery accurately.

Our validation process can include comparison of model results against historical refinery performance, operating data and expected process behaviour.

We review model assumptions, yields, capacities, constraints and economic relationships to identify areas requiring improvement.


Network LP Models

In addition to individual refinery models, LP techniques can be applied to interconnected refinery and supply networks.

Network models can evaluate relationships between:

  • Crude supply

  • Refineries

  • Intermediate products

  • Storage facilities

  • Pipelines

  • Product markets

  • Distribution systems

This allows planning decisions to be evaluated beyond the boundaries of a single refinery.


Continuous Model Improvement

Refineries are constantly changing.

New units are added, operating constraints change, product specifications evolve and market conditions fluctuate.

We therefore view LP model development as an ongoing process rather than a one-time activity.

DSCS can provide continuing support for model maintenance, modification, validation and improvement.


Training & Knowledge Transfer

Where required, our LP modelling projects can include training and knowledge transfer to enable client teams to understand, maintain and effectively use the model.


The Objective

A refinery LP model should do more than produce an optimization result.

It should provide a reliable representation of the refinery that supports better economic and operational decisions.