BLEACH PLANT PERFORMANCE ENHANCED USING
ADVANCED CONTROLS

Author

Rick Van Fleet

Company and address

Honeywell Inc., 16404 North Black Canyon Highway, MS-2S2, Phoenix, Arizona, 85053-3095, USA

email

rick.van.fleet@honeywell.com

Keywords

sequential kappa factor, multivariable predictive control, on-line kappa

ABSTRACT

Changing environmental regulations and the constant need to improve profitability were two factors in the mills' decision to apply advanced control technology to their process.  Operating in the Microsoft Windows NT environment, a configurable menu driven solution was developed which utilizes multi-layered windows providing quick, easy access to bleaching stages, regulatory controls and advanced controls.

The mills also wanted to ensure that the control system could provide the flexibility needed to handle changing process conditions and modifications. Some of the features required by the mill included automatic grade changes, automatic tuning and correcting of in-line sensors, automatic changes in production rates, and statistical monitoring of the process. 

Advanced Bleach Solution (ABS), allows for complete integration of new technologies, including on-line continuous kappa sensors, and multivariable predictive control. By continuously measuring the total bleaching load going to the process, the degree of delignification can be adjusted in order to economically spread the chemicals throughout the bleach plant. The concept of sequential kappa factor is introduced.

The package helped the mill to economically control their bleaching operations while optimizing quality control parameters and reducing environmental impact. Operating results to date will be presented.

INTRODUCTION

Over the years a great deal of effort has been focussed towards controlling the bleach plant, but the results to date have not been that impressive. Many hours of bench scale and "bucket and stop watch trials" have been performed and the "projected" results always look impressive. The real issue is that once the new schemes are placed in the hands of operators, who have not only age old ideas firmly planted in their minds on how the bleach plant should be operated, but also the constant pressure of making the required production, at the required quality level with little or no room for error, usually results in the same methods of operating taking over in a relatively short period of time. In other words, business as usual.

It has been said before that sometimes by taking a step back, you can actually get a clearer view of the path forward. This paper is about taking a look at the basic principles of the objectives of bleaching and the required measurements necessary to meet those objectives. In today's fast pace world of changing demands and changing technology, having the proper flexible yet powerful control tools at the operator's fingertips is important.

The bleach plant is the final process before the paper mill. Pulp quality such as dirt (shive content), strength (viscosity) and brightness can definitely be affected, both positively and negatively.

It is important to remember that age-old paper maker's adage of "hopefully you will give me good quality, but the key is to give me the same quality all the time". Consistent pulp quality is the key to keeping the sheet on the machine. The bleaching process is characterized by long deadtimes in the towers that range from 30 minutes to more than 3 hours. Any upset can get you in trouble for a long time. Having continuous feedback and process visibility is a necessity. Capital costs and the entire environmental permitting process make it very difficult and expensive for mills to expand their operations to include a separate bleaching line. For this reason, many mills must "campaign" the single bleach line with shorter runs of hardwood and then softwood. Minimizing the amount of transition pulp (pulp that is in between quality and fiber content), is a business necessity.

New fiber lines place a much greater emphasis on exactly how much actual delignification is done in the cooking process and how much is continued in the bleach plant. This is to improve both strength and yield. For this reason, it is ever more important to be able to accurately and reliably monitor and control the delignification process continuously.  Over application of chemical is not just costly, but all effluent streams are now closely monitored and therefore it is important to ensure that the correct dosage is applied.

Figure 1

Figure 1.0   Brightness and KAPPA through the bleaching process

The bleaching process is actually made up of two distinct functions. Bleaching is most accurately described as the process of making the pulp look whiter or brighter, but before that can happen, delignification must be completed.  Figure 1.0 shows that the rapid increase in the brightness does not occur until the majority of the lignin has been removed. The first two stages of the bleach plant are actually an extension of the delignification process that began with the cooking process. The first two stages are included in this description, since the oxidizing chemicals are not just applied in the first stage such as the case of chlorine dioxide, but further lignin removal is also accomplished by the proper use of oxygen and hydrogen peroxide. Sodium hydroxide is also very important since the removal of the chloro-lignin that is formed in the first stage is heavily dependent on the correct stoichiometric amount of caustic being applied in the second stage. This is where the concept of sequence kappa factor becomes very important. The total amount of oxidizing chemicals needed to be applied with respect to the bleaching load entering the bleach plant can be reported as the sequential kappa factor (SKf).  This definition is used to relate the total applied oxidizing chemicals to the lignin removal process, as opposed to the total applied chemical for the entire bleaching process. It is understood that some minor delignification can still be accomplished in the "brightening" stages of the process but the amount is relatively small in this context.

A way of looking at bleaching chemical demand that is not really new, is kappa factor. KF is defined as the %Total Equivalent Chlorine (TEC) applied divided by the unbleached kappa number. As mentioned, this concept has been around for some time thus the requirement of expressing the applied chemical in terms of TEC even when dealing with an ECF or 100% chlorine dioxide process. It has also been known as Active Chlorine Multiple and the Equivalent Chlorine Multiple.  Traditionally it has only been an "after the fact" calculated value since manual testing of kappa was standard. Today, not only are there batch kappa testers available but also a patented in-line continuous type (1).

Figure 2

Figure 2.0 Statistical Analysis of Kappa Factor

By calculating the applied chemical and knowing the incoming unbleached kappa number, the actual kappa factor was calculated for a bleach plant that was operating under conventional compensated brightness control. Compensated brightness is the combining of weighted signal from brightness and residual sensors located after chemical addition and before the first bleaching tower. This concept which has been popular for many years has several flaws. The most apparent one is that it does not respond directly to the true bleaching load, which is the lignin content as measured by kappa, and it is also affected by many disturbance variables.   The reaction rates of chlorine dioxide also make the residual measurement quite difficult. (2) With most mills being mandated to convert to 100% substitution, this control strategy is no longer suitable for tight control.

The feedforward measurement of total lignin, which defines the bleaching load, is the key to our strategy. This is an important deviation from compensated brightness approaches that introduce a great deal of process variability. The after mixer residual is only used when a high or low limit is exceeded. It is not used in any form of compensation of the kappa factor calculation. If a low limit were exceeded, then the kappa factor setpoint would be increased.  Feedback from the kappa measurement after the Epo stage is also used to accomplish a true delignification control. However, delignification does not occur only in the first chlorine dioxide stage. All oxidizing chemicals must be accounted for in order to both efficiently and economically spread the bleaching load out. For a DEop sequence, this value would be known as the Sequential Kappa Factor (SKf).

OXIDANT

FORMULA

FORMULA WEIGHT

ELECTRONS TRANSFERRED

EQUIVALENT WEIGHT

FACTOR TO TEC

Chlorine

Cl2

71

2

35.5

-

Sodium Hypochlorite

NaHClO3

133

2

66.5

0.53

Chlorine Dioxide

ClO2

67.35

5

13.5

2.63

Oxygen

O2

32

8

8

4.44

Hydrogen Peroxide

H2O2

34

2

17

2.09

Ozone

O3

48

6

8

4.44

Table 1.0 Oxidizing Equivalents of Different Bleaching Agents

For a typical D-Eop configuration, the sequence kappa factor would be calculated as follows:

equation 1

The benefits of controlling to sequence kappa factor are quite clear. This allows for the entire suite of chemicals to be used most effectively and also it allows for the least expensive chemicals to be maximized. This is shown in Table 2.0 both before and after control.

Table 2

Table 2.0

Sequence kappa factor is not just a series of single input single output (SISO) loops. This is by definition a multi-variable control problem with many interactions. For example, if the unbleached kappa number entering the bleach plant suddenly drops, and the only control action that is made is to lower the chlorine dioxide, then after a delay equal to the retention time of the first two towers, the CEK number will drop considerably if no action is taken on the oxygen or hydrogen peroxide. If this is not the case, then the past dosage rates of these chemicals were incorrect and being under utilized. In this case the kappa number is acting as the disturbance variable.

What this means in simple terms is that as the disturbance variable, DV (unbleached kappa) changes, the controller makes co-ordinated changes to not only the chlorine dioxide, but also to the oxygen and peroxide. Included in this matrix is also the sodium hydroxide required for proper pH control and extraction. The manipulated variables (MV's) are the chemicals; chlorine dioxide, oxygen, peroxide and sodium hydroxide. The controlled variables would be the CEK or degree of delignification, the pH and the pre-tower residual. A key factor here is that absolute setpoint control is not required for optimal performance. It is dynamic and the concept of a range control algorithm is introduced. This means that there is flexibility in the solution of the control problem and that the tuning can be adjusted to allow for a solution based on the degree of robustness required. When the cost is also involved, based on the reaction rates, bleaching efficiencies etc. several solutions can exist.

Many control packages only address the first layer or two of the control hierarchy. It is not acceptable to only meet the first level objectives of a 'single loop' controller. Just because the flow meter says that the setpoint is equal to the process variable, this does not indicate if you are meeting your business objectives. Today it is important to be able to manage your process in a more coordinated, real-time fashion.  By utilizing the latest technologies of Microsoft's NT and networking capabilities, data can more easily be transformed to information and sent to anyone that requires it. This includes inventory and cost reports to accounting, production reports to scheduling, quality data to the technical department for example. This also includes giving management a real-time view of exactly what is going on.

Advanced Bleaching Solution (ABS) is designed to fulfill many more requirements than simply manipulating some chemical valves. By utilizing the control technology of multi-variable predictive control (MPC), the bleaching process is treated as a single control entity rather than just a collection of single input, single output loops. The design allows for the real-time consumption of chemicals to be linked to the mills MIS system that traditionally had separate calculations of the cost of bleaching. This allows for real-time cost calculations that can be trended and tracked. The package gives a coordinated view of the entire fiiberline that shows both upstream and downstream inventories. By treating the entire bleach plant as a single entity, rather than a collection of stages, we are able to truly optimize the process, dynamically. This means that cost can now be used as a constraint, and by shifting or spreading the bleaching load we can economically optimize the process while maintaining the pulp quality. Through the connectivity to the plant wide information system that now exists, it is also possible to have current real-time costing and inventory data that gives a better report on bleaching costs. This sharing off information is also important when you begin to view the bleach plant as an integral part of the entire fiberline. This allows for a coordinated ramping of production or changing of grades.

If the pulp coming into the bleach plant were perfectly uniform there would not be a need for an advanced control system. Sources of variability in an operating environment include black liquor carryover, consistency fluctuations, species changes, instrumentation problems, production rate changes, pH changes and instrumentation problems. This list is just some of the most common sources of variability that any control system must contend with. Many of these are not even measured and thus are known as unmeasured disturbance variables. They are all present and must be accounted for.  Some such as instrumentation problems can be fixed after conducting an audit. It is important to recognize that there is value in fixing instrumentation problems. Training can help with tester error and operator problems. Often a mill looks at monthly or running averages and claims that they do not have variability problems. It is this variability that the control system must address. Shifting or changing operating targets is all that much easier to do and economical as well as quality optimization can be accomplished.

CONCLUSION

Better technology leads to lower project implementation costs and higher benefits. Multi -variable predictive control coupled with on-line kappa sensors provides control and economic justification of processes that have significant interaction between variables. The real value of this approach is that it allows you to consider the entire bleaching process as a single economic and process entity rather than a collection of independent and isolated loops. Today Windows based tools exist and allow mills to keep the process within operational constraints while optimizing performance measures, such as cost per ton. Sequential kappa factor is a powerful way to economically shift the bleaching cost.

REFERENCES

1. Millar, O., Van Fleet R. Continuous In-Line Kappa Measurement System (U.S Patent No. 5 ,953,111).

2. Williamson, M., Pulp & Paper Vol. 73(7) pg. 37

BACK TO TOP

[Home] [Title] [Author] [Organisation] [Keywords]