Consultant Canfor Ltd.

Erin, Ontario Prince George, B.C.

Scott Fairless Todd Johnson

Neles Automation Inc. Canfor Ltd.

Vancouver, B.C. Prince George, B.C

In October 1999, Canfor Ltd.’s, Prince George, B.C. upgraded the control strategy from Compensated Brightness Control (CBC) to Compensated Kappa Factor Control (CKFC) in the Do stage of their market pulp mill. The pulp mill bleaches oxygen delignified softwood at 20 kappa no. using a DoEopDEpD sequence. The pulp mill produces approximately 290,000 ADt per year at 90% ISO brightness for the international market.

An assessment of the Compensated Kappa Factor Control (CKFC) strategy compared to the prior Compensated Brightness Control (CBC) strategy over both long and short term operating periods showed:

The operators readily adopted the new control strategy and the control system has averaged 97% availability.

Canfor Ltd., Prince George B.C., produces 290,000 ADt per year of fully bleached oxygen delignified softwood pulp at 90% ISO brightness for the international market. The softwood is comprised mostly of spruce, fir, and pine.

The fiberline consists of a single vessel hydraulic digester, pressure diffuser, 2-stage pressurized knotters, followed by two 13’-6"x 22’ vacuum washers discharging to a single 260 ton high density storage tank. Brown screens with stingray rotors and barrier screens feed the Pre-Oxygen Delignification washer followed by a single vessel O₂-Delig reactor and two twin-screw wash presses. The medium consistency stock is then fed to a single 200-ton high density storage tank. Approximately 80% of the discharge of the high density passes through a 14’x28’ drum washer. The bleach plant is conventional five–stage design utilizing a 13’-6"x22’ drum washer after the first bleach stage, followed by 11’-6"x22’ drum washers for the remaining bleach stages. In mid 1980s, the initial CEDED bleaching sequence was upgraded to a CdEoDED sequence, and converted to Elemental Chlorine Free (ECF) in 1994. The bleach plant utilizes MC pump technology

The primary driving force for implementing Compensated Kappa Factor Control (CKFC) was to improve process stability and lower the overall bleaching cost.

Prior to Compensated Kappa Factor Control (CKFC), Compensated Brightness Control (CBC) was used to control ClO₂ addition to the D₀ stage. This control method used Kajaani Cormec brightness and Polarox residual sensor measurements located just after ClO₂ addition to calculate a compensated brightness value. The sensors provide brightness and residual feedback control. Operators adjusted the compensated brightness value, based on their experience with incoming pulp characteristics (i.e., Kappa number, carryover, etc.) to adjust ClO₂ addition to achieve target brightness and delignification.

Sodium hydroxide addition to the E_0P stage is controlled by one of two methods of supervisory control:

• operator set point of percent NaOH on pulp (normal operation); or
• pH control using a sensor located at the inlet of the Eop J-tube.

Hydrogen peroxide and oxygen addition to the E_0P stage are controlled by percent on pulp, (operator selected set point).

The D₁ stage ClO₂ addition is controlled in a similar fashion as the Do stage using a compensated brightness set point. However, in this case, the brightness sensor is located before ClO₂ is added. This is done to keep the sensor free of material that precipitates on the sensor as the pH changes after ClO₂ addition. A residual sensor is located near the base of the D₁ tower, downstream of the ClO₂ addition point.

There are two control strategies in this stage:

added based on percent on pulp as a backup mode of operation; and

• Hydrogen peroxide is added based on a compensated brightness set point. The brightness and residual sensors are located after NaOH addition.

D₂ stage ClO₂ addition is controlled by percent on pulp.

In the modified strategy, the D₀ stage is controlled by Compensated Kappa Factor Control. This control method incorporates incoming bleach plant Kappa number (which is measured by a Kajaani Kappa Analyser) with brightness and residual measurement just after ClO₂ addition. With CKFC, ClO₂ dosage stability problems are minimized by combining the kappa number and the brightness and residual sensor outputs into a single measurement. CKFC takes into account changes in incoming kappa number. The ClO₂ addition is adjusted according to kappa number and corrected by using brightness and residual measurements installed after ClO₂ injection. The control algorithm uses the incoming kappa number as a form of feed-forward control and the brightness and residual probes as a means of a dynamic feedback.

The Cd stage was modified to high ClO₂ substitution in the mid 1980s, which resulted in a designed retention time in the piping after chlorine dioxide addition. During ECF operation, to ensure that this transition piping from the ClO₂ injection to the brightness and residual probes acts as a constant reaction vessel, a dual feedback pH loop with feed forward was implemented. Dilute sulfuric acid is metered and regulated (flow control) at the suction of the first stage feed pump with an immediate feedback pH sensor located just prior to the ClO₂ injection. This inner pH loop incorporates a calculated feed forward based on the ClO₂ injected, with long term correction via a final pH loop measured at the inlet of the upflow tower. Good pH control is essential with the brightness and residual probes in their current location to ensure that the reaction rate is stable.

No changes were made to the existing E_0P, D₁, E_P, and D₂ stage control strategies.

The operations displays were constructed to give as much pertinent information as possible, while retaining a simplistic approach. On the display, the operator has the ability to choose the new CKFC as well as having two backup modes of operation, the first being the original compensated brightness control (CBC), and the second being simple dosage control. While running the new CKFC strategy the operator can readily bias the dosage by increasing or decreasing the kappa factor algorithm. This simple function adds sufficient flexibility as to keep the controls in operation throughout all abnormal operating conditions (except for equipment malfunction).

The successful startup was gauged by the fact that once the controls were placed in supervisory, they remained on supervisory control throughout the tuning phase. There has only been one brief period (8 days) when the controls were not utilized due to an optical lens problem on the analyzer. Today the operations personnel rely heavily on the control strategy and the process information gained from the kappa analyzer.

The most obvious benefit of the CKFC control strategy as compared to the CBC strategy, is the prevention of over-bleaching during upset conditions. The control strategy allows the ClO₂ dosage to decrease significantly with low incoming kappa pulps, yet allows the D₀E_op Kappa number to float upward with high kappa pulps. The latter scenario introduces a slightly higher DoEop kappa number Coefficient of Variability (COV) into the D₁ stage (during upsets), yet reduces the overall bleaching chemical consumption. In order to compensate for the floating kappa number into the D₁ stage, the tuning of D₁ compensated brightness controller was adjusted to be slightly more aggressive. As a result, the D₁ brightness COV and the total kappa factor COV decreased with the added stability.

The data is compared to the average of the seven oxygen delignified mills in Canada with the lowest (Low) overall oxidation demand per kappa number and the average of the seven oxygen delignified mills with the highest (High) overall oxidation demand per kappa number [1].

The first stage kappa factor decreased from 0.19 to 0.17 and is now among the lowest in Canada. The improved oxidizing chemical consumption was sustained over the first two stages and again is at the levels of the lowest in Canada.

Sodium hydroxide increased unexpectedly. Furthermore, analysis of the data show that the same end pH was achieved in all operating periods. Interestingly, the ratio of % NaOH on pulp in the Eop stage to the % equivalent chlorine applied in the Do stage for the CKFC periods is the same as the average of all oxygen delignified mills in Canada. This brings into question the % NaOH data for the reference CBC period. Therefore, two assessments are made on the impact on bleaching cost, one based on all bleaching chemicals including NaOH, and the other based on oxidizing chemicals only.

Overall bleaching oxidizing chemical consumption also decreased as shown by the sequence kappa factor. The sequence kappa factor decreased approximately 8-9% and this lowered bleaching costs by approximately 2% when compared directly and 6% when compared for oxidizing chemicals only. The benefit is likely even greater as the CKFC periods produced pulps with 1% ISO higher brightness than the CBC reference period (Table 1).

The benefit improves when compared at the same incoming kappa no. The bleaching costs are lowered by approximately 5% when compared directly, and 10% when compared for oxidizing chemicals only. The benefit is likely even greater as the CKFC periods produced pulps with 1% ISO higher brightness than the CBC reference period (Table 1).

* Oxidizing chemicals only, ** prorated to 20 Kappa No., *** oxidizing chemicals only and prorated to 20 Kappa No.

This analysis compares Compensated Kappa Factor Control (CKFC) to Compensated Brightness Control (CBC) as before, however in this case, the data compares a week with the control on to a week with control off and then a week with control on. As before the data was obtained from the mill data management system and is organized into the following periods:

As Figure 9 clearly shows, the Do kappa factor increased from ~ 0.16 to 0.18 when CKFC was turned off and returned to a lower value when CKFC was turned back on.

As in the case of the Do kappa factor, the DoEop kappa factor increased when CKFC was turned off and returned to the lower values when CKFC was turned back on. The data indicates that the operators used a slightly more conservative control strategy when using the CBC strategy.

As expected, there is little difference in the NaOH:Equivalent Chlorine ratio comparing the different periods of operation. These ratios are similar to the average values found in Canadian bleached kraft pulp mills [1] and provides additional support to the possibility that the NaOH:Equivalent Chlorine ratio in the long term reference period may be in error.

The overall oxidizing chemical demand per unit kappa number (sequence kappa factor) increased ~ 6% when CKFC was turned off and returned to the former value when turned back on. This is reflected in a 5% increase in bleaching cost as shown below. The bleaching cost increased 7% when compared at the same kappa number as shown in Figure 13.

The authors gratefully acknowledge the support of the Canfor Ltd., and Neles Automation Inc. Most particularly we wish to acknowledge the bleach plant operators. Their contribution helped to make this project a success.