[Home] [Journal papers]

peer reviewed logo


Jukka Puhakka and Ossi Tolonen,
Metso Automation Ltd

Presented at the 31st Pulp and Paper Annual Meeting, ABTCP, Sao Paulo - SP - Brazil

Chemical recovery and causticizing process

Chemical recovery is an important part of the kraft process: recovering chemicals from the spent cooking liquor, reconstituting them to form fresh cooking liquor, producing energy from the incineration of organic residuals and minimizing air and water pollution (fig. 1).

Online alkali fig 1

Figure 1. Principle of causticizing

Causticizing plays an important role in the chemical balance of a kraft mill, because this process regenerates cooking liquor. The object of the causticizing process is to turn inactive sodium carbonate (Na2CO3) into the active cooking chemical, sodium hydroxide, and of course to make sure that the conversion efficiency of carbonate into hydroxide is as high as possible. The process can be divided into three parts: slaking, causticizing and white liquor clarification/filtration, fig. 2.


In slaking, lime (calcium oxide CaO) reacts with the hot water (H2O) mixed in the green liquor, generating calcium hydroxide Ca(OH)2 and heat (about 1kJ/kg CaO).

H2O+CaO- => Ca(OH)2 + heat

Online alkali fig 2

Figure 2. Causticizing process.

The slaking reaction begins and proceeds only when the green liquor temperature is +60C (140F) or higher. However, the slaker temperature should be kept low enough to prevent it from boiling but still high enough to maintain maximum reaction speed. Normally the temperature of incoming green liquor is between 85-91C, and the temperature increases in the slaker by 14-15C. This slaking reaction indicates the quality of the lime, but the lime feed to causticizing cannot be controlled based only on the slaker temperature difference. As an example, if lime reactivity goes down (i.e. very hard burnt lime), it takes longer to slake and the lime dosage must be increased to achieve the target causticizing degree. Unreacted excess lime must then be removed from the process through the slaker's classifier part, and again the need for fresh lime increases and easily causes overliming. On the other hand, if lime reactivity goes up, the slaker temperature will also rise and lime feed decreases; this means a lower causticizing degree than expected.


In this endothermic causticizing reaction, calcium hydroxide (slake lime) Ca(OH)2 reacts with the sodium carbonate (Na2CO3) in the green liquor. This reaction generates sodium hydroxide (NaOH) and calcium carbonate (lime mud) CaCO3.

Ca(OH)2 + Na2CO3 => 2 NaOH + CaCO3

Causticizing degree (CD%) describes the completeness of the reaction.

CD% = NaOH/NaOH + Na2CO3 * 100

High conversion efficiency is desired to reduce the load of inert sodium carbonate in the recovery cycle. Typically the causticizing degree is around 80-90%, depending on alkali concentration and sulfidity level. Goodwin's curve (fig. 3) describes the highest achievable causticizing degree. In many mills operating without control, the causticizing degree is far below the theoretical value in order to avoid overliming. Fig. 4 shows a schematic description of the causticizing process, and fig. 5 contains actual data on causticizing degree from European and US mills. In fig. 6 the causticizing degree is also described as a function of total titratable alkali, active alkali and sodium carbonate in white liquor.

Online alkali fig 3

Figure 3. Goodwin's Curve

Online alkali fig 4

Figure 4. Schematic description of causticizing

Online alkali fig 5

Figure 5. Data on causticizing degree


White Liquor Clarification / Filtration

Lime milk from the causticizing vessels contains:

- water (about 74 weight %)

- various alkalis (about 12%)

- lime mud (about 14 %)

After causticizing the lime mud is separated from the water and dissolved alkalis to produce white liquor. Lime mud is removed mechanically, either by sedimentation or nowadays commonly by filtering (pressure/vacuum). Good efficiency of the white liquor process (particularly filtering) requires careful control of the lime milk composition. If overliming occurs , the filters will require more service and may even get blocked. The target lime mud content in white liquor should be less than 30 mg/l.



Kajaani Alkali Analyzer is a fully automatic, on-line sampling and titration analyzer for green liquor, causticizing and white liquor, made by Metso Automation Kajaani Ltd. Using a field proven on-line sampling system and autotitrator, it provides outputs for control, based on actual process chemistry titration results.

The Alkali Analyzer system consists of sample unit, with 4 or 8 sample points, and a measurement unit with one or two titration modules (fig. 2).

In a typical application, samples are taken from green liquor, causticizing vessels and white liquor to digesters. The analyzer takes samples and analyzes them automatically. The analysis is based on the ABC-titration procedure, the most widely used method in kraft mill laboratories. Fig. 7 presents the measurement principle and sequence. The analyzer measures the absolute values of Sodium Hydroxide (NaOH), Sodium Sulfide (Na2S), Sodium Carbonate (Na2CO3) and calculates Effective Alkali (EA), Active Alkali (AA), Total Titratable Alkali (TTA), Causticizing Degree (CD%) and Sulfidity (S%). The results can be monitored from the analyzer and sent to the mill DCS system as analog or digital signals.

Online alkali fig 6

Figure 6. Causticizing degree

Online alkali fig 7

Figure 7. Alkali Analyzer's measuring sequence



The Kajaani causticizing control incorporates a Kajaani Alkali Analyzer and Metso Slaker Control. Its main target is high and stable causticizing degree, which in practice means homogeneous quality, high strength of white liquor, higher production capacity, and reduced operating costs in the causticizing plant and in the rest of the pulp mill. The input measurements of the control are: green liquor temperature, density and flow, slaker and lime milk temperature, reburnt lime feed and alkali strengths (fig. 8). The density measurement is temperature compensated. The reburnt lime feed is a function of conveyor momentum and thus a function of conveyor motor current. When possible, the massbalance can also be used to calculate the reburnt limefeed.

Online alkali fig 8

Figure 8. Principle of causticizing control

The control generates several outputs: green liquor TTA control with density, causticizing degree prediction, correction of the liquor/reburnt lime ratio, and the quality of lime milk, e.g. possible overburning of lime or overliming.

Green Liquor TTA Control

A model, converting green liquor TTA to density, is made and constantly updated by the Alkali Analyzer's analysis and density measurement. Small variations in sulfidity and other sodium based components are detected for better green liquor TTA control. When using this model, the operator sets the TTA value instead of density. With the model a constant TTA can also be calculated from the density using the formula backwards.

Causticizing Degree Prediction

Causticizing degree is predicted from green liquor temperature, density and flow, using the Alkali Analyzer's measurements and the model presented earlier. With sulfidity and the calculated constant TTA, the theoretical maximum of causticizing degree is calculated according to Goodwin's curve.

Slaker Control

Green liquor flow is used as a feed forward signal for slaker control and CD% prediction. CD% set point after the slaker is set according to the flow rat, also taking care of reaction time changes. Lime milk CD% after the slaker is used as a fast feedback signal for temperature difference control, correcting the set point by the difference between CD% set point and analysis. CD% measured by the analyzer after the last causticizing vessel forms a slow feedback loop.

The actual CD% which indicates the quality of reburnt lime and slaker status, is compared to the setpoint given by operator. CD% set point after the slaker is then corrected with a function of the difference. When this, too, is connected to set the level for lime feed, it makes the controller self-tuning.

Follow-up and Reporting

The outputs and their effect on the system are reported as numeric data, as trends and/or bars, either in base level displays or in supervisory level displays. Causticizing degrees are shown as numeric values next to the process picture with the other analyzer information. The predicted CD% signal is delayed and shown in the same trend with the analyzer's results. The predicted CD% is a continuous signal, the others are discrete. Correction to the lime/liquor ratio is shown either as percentage or as absolute value.


The main display shows the latest measurement results and calculated results from Alkali Analyzer with time labels.

The tuning and diagnostics display contains important information of the analyzer operation and the progress of titration. In the tuning section, the operator change the sampling order simply by typing a new order in the sampling list. The status of sampling and the latest alarms are presented, as well as some self-diagnostic information from the analyzer.


The control system described above is currently in use in a Finnish pulp mill. As the trial is still going on at the time of writing, process data cannot be included in this paper. Results from the mill will be presented in the conference.


Several benefits will be achieved with on-line alkali strength measurement and slaker control. Fully automatic, on-line sampling and titration of green liquor, causticized liquor and white liquor help to stabilize the TTA of green liquor and to automate slaker control, enabling the optimization of the causticizing process.

This results in lower emissions, fewer process bottlenecks, higher production capacity and reduced operating expenses in the caustic plant and in the rest of the pulp mill.

Authors' contact details:

Jukka Puhakka, Product Manager, Metso Automation Inc., P O Box 237, FIN-33101 Tampere, Finland.

Ossi Tolonen, Product Manager, Metso Automation Kajaani Ltd., P O Box 177, FIN-87101 Kajaani, Finland.


[Home] [Journal papers]