Varo mill optimizes lime kiln with predictive control: the Sodra Cell mill in Varo, Sweden is using multi-variable model predictive control to optimize lime kiln process and cuts costs.

Pulp and paper companies face the challenge of being highly capital intensive while serving very mature, slow growing markets. It is critical that companies understand the process capabilities of their entire operation in order to select the technology capital projects that will eliminate productivity bottlenecks, improve quality, and reduce costs.

[ILLUSTRATION OMITTED]

The case study in this article is an excellent example of utilizing advanced process technology to achieve the "golden triad" of:

* improved productivity

* improved quality

* reduced costs with a fast pay back time.

BACKGROUND

During the spring of 2004, Sodra Cell's Varo, Sweden mill implemented multi-variable control on their lime kiln (Figures 1 and 2). For the control solution, the mill selected Invensys Process Systems (IPS), which developed Connoisseur products for the optimization of all pulp mill areas. The primary justification for the investment was to use advanced control technology to raise the production capacity of the lime kiln. The lime kiln had become a limiting factor for the mill's pulp total pulp production, especially during long periods at high production rates.

Another key goal for the advanced control system included producing reburned lime with consistent quality at the lowest possible cost, i.e. minimizing energy costs. The results of the optimization are as follows:

* Lime Kiln capacity was raised by 5.3% calculated as the amount of CaO produced.

* Lime quality is more consistent, and the variability in residual carbonate was reduced by 74.8%.

* Energy usage was reduced by 16.38 kg oil per Metric Ton (MT) of CaO.

MULTI-VARIABLE MODEL PREDICTIVE CONTROL

The new Invensys multi-variable controller has no practical limitations with respect to controlled and manipulated variables. The controller has features for: optimization, fuzzy logic control, inferential control, adaptive control and neural networks. The Model Predictive Controller has internal model identification capability and dynamic updating which simplifies control model tuning, as identified using PRBS (pseudo random binary sequence) testing.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

The MVMPC Solution includes the following manipulated variables (MV), feed forward variables (FV) and controlled variables (CV):

 1. Mud flow to mud filter                     MV
 2. Mud density to mud filter                  FV
 3. Product/Consumer gas fuel flow             MV
 4. Oil fuel flow                              MV
 5. Induced draft fan speed                    MV
 6. Damper position                            MV
 7. Kiln hot end temperature                   CV
 8. Kiln cold (feed) end temperature           CV
 9. Flue gas temperature after mud dryer       CV
10. Flue gas 02 content                        CV
11. Reburned lime residual carbonate lab test  Fuzzy Logic Signal

STABILIZING CONTROL

A requirement for improved performance is the ability to stabilize kiln operating conditions. With the kiln being heated with two different fuels--producer gas and oil--it is critical to identify the impact of each fuel on the calcining process. Product gas is generated from burning bark and has priority due to the low cost of the fuel. The control strategy is therefore to burn as much producer gas as possible, and only burn oil when there is not enough producer gas.

The kiln hot end and cold end temperatures are controlled to achieve consistent calcining with the lowest possible energy usage. The temperature curves (Figures 3 and 4) show how control has improved over time, becoming more stable. The upper trend shows the (result) testing period compared with the lower trend which shows a selected period of conventional process control with good control at high load.

Oil flow is controlled to supplement the producer gas (Figure 5). This control has to be responsive and stable to compensate for fluctuations in producer gas flow, including occasional production shutdowns. With this feature being well tuned, large amounts of energy can be saved while maintaining the requisite calcining temperature profile.

Results as a reduction in variability (1-Sigma) are as follows:

* Hot end temperature stabilization: 45.9%

* Cold end temperature stabilization: 80.9%

* Flue gas temperature stabilization: 88.2%

* Reburned lime residual carbonate: 74.8%

* 02 content in flue gas stabilization: 78.6%

SUMMARY

Sodra Cell Varo used advanced process control to raise its lime kiln capacity, reduce energy usage, and improve the overall quality of the lime. The investment paid for itself in less than 12 months, while the productivity, quality and cost benefits will continue for years after the installation. Installing lime kiln advanced control system empowered the operations staff to focus on keeping the mill running at target production and stop worrying about balancing kiln operation costs, lime quality, and production parameters. Advanced technology can help operations staff focus the entire operation instead of just keeping the one element running.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

Many technology projects/solutions can help pulp and paper companies achieve the "golden triad" of improvements in productivity, quality, and costs. The key to success in these projects is understanding process capabilities and maintaining the resources to search out these opportunities.

[ILLUSTRATION OMITTED]

TED MCDERMOTT, SOLUTIONS! EDITORIAL BOARD

IN THIS ARTICLE YOU WILL LEARN:

* How the Sodra Cell mill in Varo, Sweden is using MVMPC to optimize lime kiln process and cuts costs.

* The results of the optimization.

ADDITIONAL RESOURCES

* "The Promise of technology: Past, present and future," by Ted McDermott, Solutions!, June 2004. To access this article, enter the following Product Code in the search field on www.tappi.org: 04JUNSO51. Or call TAPPI Member Connection at 1 800 332-8686 (US); 1 800 446-9431 (Canada); [phone] (International).

* 2004 TAPPI Paper Summit, Papermaking, PCE & I, and P & PO & Environmental Conference Proceedings CD-ROM. To order this disc, see information in item above. Product Code: SUMMIT-04 ISBN: 1595100350; Member Price: $55.00, Non-Member Price: $80.00.

Item  Parameter               Data         Units

1     Kiln diameter                 3.3    Meter
2     Kiln length                 113      Meter
3     Lime Production             320      MTon/day
4     Days in Operation           350      Days/year
5     Annual Pulp Production  390 000      ADMT/year

Table 1: Operational data