WHAT ARE THE CHALLENGES THAT ARE BEING FACED IN THE POLYMER INDUSTRY?

In polymer manufacture1st Quality has been accepted as a product that is within a +/-10% tolerance band. For the polymer processor the challenge is to convert the highest proportion of raw materials into saleable product. Large batch-to batch variation in raw materials equates to losses in scrap whilst the downstream processor makes adjustments to the equipment. Wide variations in successive batches will incur greater use of raw materials to meet strength tests etc. The polymer processor therefore creates an ongoing demand for tighter Q1 tolerances, which is currently of the order +/- 5%. This level is difficult to achieve with most control technologies, leaving the polymer manufacturer with the ultimate challenge of upgrading the polymerisation control system.

For the polymer manufacturer, the operational challenge is to fit in any process upgrade with minimal impingement on production. A typical polymer plant producing a number of grades of material presents challenges in fine process control. The first is to make a safe transition from the end of a product campaign and to hit the new target grade in the minimum time. The second is to make the in-grade product consistently throughout each campaign. Product changeover and other out of specification running are periods of 2nd quality. These materials are generally sold at a loss.

Molecular Control Systems

As the provider of the control solution, the challenges to Molecular Control Systems (MCS) are numerous. The prime challenge is to maintain strict project confidentiality so that sensitive polymer manufacturer information is not leaked in any way. MCS and the polymer manufacturer are both to be contractually obliged to restrict the use of the carefully engineered control solution to the designated plant. The solution is unique to each plant, making replication elsewhere ill advised. The project challenges include information exchange, project costing, planning, executing extreme data intensive model building work in a realistic time scale, introduction of new software procedures into existing computer facilities, documentation and training. The engineering challenge is to deliver the control scheme that matches the manufacturers financial case. The special challenge is to enhance the team awareness of QC and production.

HOW IS THE PROCESS BROUGHT INTO BETTER CONTROL?

The design intention is to equip the plant, as necessary, with state-of-the-art analysers, in order to obtain accurate and timely measurement of key process data. The equipment list includes Gas Composition and Melt Flow Index analysers. The object is to progressively upgrade the control system of the polymerisation process. The most cost effective way for all concerned is to work in a two-sided consultative manner. At the outset project objective and delivery are to be agreed along with the targets of budget, time scale and benefit.

The project is to be carried out in three stages:

Stage 1 – closed loop control – for mid-campaign periods of steady running

Stage 2 – transition control – for grade changes

Stage 3 – process optimisation – progressive refinement of the new control system

STAGE 1 – Closed Loop Control

  • Review control capability of plant
  • Install/upgrade any necessary analysers and process equipment
  • Build closed loop control

This stage firmly establishes operations in the frame work of a well-equipped and well instrumented plant. The full time availability of high quality analyser data ensures correct build-up of the closed loop control algorithm. In these circumstances the control system delivers a more repeatable structure for each polymer grade. Narrow grade targets and pass bands become readily achievable.

This section of the project looks after a lot of the product quality issues but the process will still need to be manually guided in the transition periods until the completion of the next stage. However, the collection of high quality process data will have already commenced.

STAGE 2 – Transition Control

  • Collect process data of all transitions
  • Build models of grade transition
  • Implement transition control with updated models

In this part of the programme we progressively overcome two major control challenges – making the transitions quickly as possible and hitting the new target grade on transition exit every time.

Analyser delay mandates that feed forward control is to be used in the regulation of transition materials. Each transitions requires its own particular change-over sequence and may contain more than one intermediate grade to accommodate any large change in polymer structure. The programme is to build each grade-to-grade transition from plant data. This requires the use of an advanced computational facility. It is vital to the polymer manufacturer that all confidential information is processed in a secure environment so that the competitive position is not compromised. Loughborough University operate such a facility and has made this available to MCS and the polymer manufacturer under confidentiality. In the Loughborough computer unit, we can validate these models before committing to the final implementation in the production environment. This potentially turns a lengthy manually operated step-check-step-check series of events into a safer procedure that avoids using up a lot of valuable production time.

STAGE 3 – Optimisation

This programme builds on the closed loop and transition work. As more data comes in it is possible to progressively tighten the changeover time as well as improving the accuracy of entry into the closed loop part of the campaign. Optimisation progressively improves the ‘bottom line’ of the operation in the reduction of energy and consumption of raw materials. A case study has shown the ROI of the total project to be in the region of 6-9months. This project can help product development, as new variations in product formulation can now be better realised in the environment of fine control of the process. Fine control also reveals the why’s and wherefores of other possibilities in process improvements and de-bottlenecking.

  • Start with normal grade schedule to reduce permutations and degree of change
  • Use existing models, data and sequences as starting position
  • Test and refine models in off-line computer facility
  • Start with slow transition sequence and improve
  • Last but not least, the project is a collaborative contract.