LIME KILN OPTIMISATION: MANAGING THE INPUTS
TO STABILISE THE OUTCOME

Heikki Imeläinen, Mauri Loukiala and Urpo Launonen

Presented at the 59th Appita Conference, Auckland, New Zealand 16-19 May 2005

ABSTRACT

This technical paper discusses the requirements for optimized lime kiln control and presents an optimization control package which has been applied to a number of lime kiln processes. These high level supervisory systems oversee the coordination of the basic lime mud washing, fuel combustion, lime burning and kiln operation controls to stabilize the process and to attain the even lime quality required for a more stable causticizing process. These optimizing controls achieve a steadiness of operation that the most experienced operators would find difficult or impossible to achieve.

To achieve this steadiness of operation, these optimization control systems strike a balance between input and output conditions and the internal kiln temperature profile by employing process running models, adapted from operating experiences, and fuzzy logic controls. Once stabilized, the process energy consumption can be reduced. Also, under these conditions, gaseous effluents can be minimized.

Results from two European applications of lime kiln control are presented. In addition to achieving a high degree of stability in the process and in the lime quality these controls have resulted in significant energy savings and have increased the capacity of the causticizing process.

KEYWORDS

Lime kiln optimization, controls, fuzzy logic

INTRODUCTION

The lime kiln operation in a kraft pulp mill is one of the most challenging to control. First of all, it is a complicated mixed-phase chemical reaction process. Quickly moving combustion gases provide the heat input to chemically change the slowly moving solid-phase lime mud into lime.

Secondly, the many different types and degrees of process delays are very significant and difficult to manage manually. Adequate, but not optimum, manual control takes a well developed feel for the process and careful manual management of the individual control loop targets. In a process with such long time delays, it is common for operators to overreact to changes, resulting in long periods of instability and variable lime quality. The result of a change in operation is often seen by the operators in the next 8-hour shift. And any attempt at optimizing energy consumption is futile if the process is not stable.

The challenges to the operators of a lime kiln are considerable. The availability of the lime kiln can be reduced significantly by the buildup of rings which plug the kiln and need to be physically removed. Availability is also reduced by overheating of the refractory bricks, causing the bricks to crumble and fall.

Specific energy consumption (GigaJoules / ton of CaO) is increasingly an issue, as energy costs are rising rapidly and the emission of greenhouse gases needs to be reduced. Variable lime mud quality (residual carbonate) can cause instability in the causticizing process, resulting in overliming and unstable white liquor quality.

OVERVIEW OF OPTIMIZATION CONTROLS

To address these lime kiln process stability problems Metso Automation has installed 13 DNAlime optimization control systems in kraft pulp mills. These high level supervisory systems oversee the coordination of the basic lime mud washing, fuel combustion, lime burning and kiln operation controls to stabilize the process and to provide even lime quality for a more stable causticizing process.

The hierarchy of optimization controls of a lime kiln process is shown in Figure 1. Supervisory-level controls determined the setpoints for lower level process loop controls. Fuzzy logic is employed in some cases to provide more consistent control when input conditions are changing.

Figure 1: Optimization controls supervise the basic lime mud washing and lime burning operations. Running models, including fuzzy logic, determine the best operating parameters.

Figure 2 shows how the recipes define the controlled variables of the lime mud filter and the kiln. Running models (or recipes) supervise the entire operation. Production rate controls ensure stable steady-state operation and smooth transitions during production rate changes by coordinating lime mud density and lime mud flows to a production rate target.

Lime mud filter controls ensure a clean lime mud with evenly high dry solids content. The lime kiln controls primarily affect the energy input (burning power) and the temperature profile in the kiln. The flue gas fan rpm is controlled to achieve the correct kiln temperature profile balance between the feed end and the burning end and the correct oxygen level during burning.

Figure 2: Running models, or recipes, determine the targets for lime mud filter and kiln controlled variables.

LIME MUD FILTER CONTROLS

The production rate control, calculated from the dry solids flow to the mud filter, supervises the operation of the lime mud filter and the kiln, to make transitions with minimum upsets. Filter vat level control stabilizes the formation of mud cake on the wire. Washing water control ensures even dry solids content and low impurity level in the mud cake. The two objectives are often conflicting since higher dry solids can mean more impurities. In the control, the best combination of these two objectives is optimized. Low impurity levels in the mud ensure lower TRS emissions.

The filter drum speed is adjusted according to the production rate target, which is compensated by fuzzy logic controls to achieve long term stability in lime mud dry solids. The dry solids calculation can even detect a malfunction in a lime mud dry solids meter. In case the meter malfunctions, an energy balance calculation determines the energy required for water evaporation, hence the amount of water evaporated.

Lime mud filter controls are shown in Figure 3.

Fuzzy logic modifies the recipe values ( set points) in the lime kiln. The normal recipe value represents the optimal set point for normal process conditions. Fuzzy logic adapts the set point for actual current process conditions.

Figure 3: Line mud filter controls

KILN CONTROLS: A BALANCING ACT

In the kiln, the temperature profile from the inlet to the outlet is the single most important variable, which, if properly controlled, will result in even lime quality and a lower tendency to form rings, which can cause significant process downtime. The calcination process is stabilized by controlling the temperature profile in the lime mud as it progresses through the kiln. The temperature of the lime mud cannot be practically measured so the temperature profile of the flue gases is controlled. This affects the lime mud temperature and therefore the calcination process. By stabilizing the process, the starting point of reaction zone is fixed . Since the calcination process is a bi-directional, temperature-induced chemical reaction, any reversal of the reaction back to CaC03 as a result of unstable temperature gradients is avoided. The temperature targets are not constant, but are continually optimized according to ring size in the kiln, lime mud dry solids, causticizing efficiency, etc.

The flue gas temperature profile is managed by:

• The fuel feed, which regulates the temperature level.
• The primary air flow, which affects the flame shape and provides effective heat transfer.
• The flue gas fan rpm, which affects the kiln temperature profile. Oxygen and carbon monoxide measurements modify the draft flow control to maintain a proper level of excess oxygen in the flue gas and low carbon monoxide emissions.
• The kiln rpm, which affects the lime mud depth (degree of filling).

The integral of the profile is controlled by burning energy (fuel flow control). If the feed end is cold and burning end equally hot, then the integral is correct. The profile is correctly maintained only by changing the flue gas fan speed. There is no need to change the fuel flow.

The final carbonate level in the burnt lime modifies the burning end temperature. The profile control then responds to this change in burning end temperature.

The kiln burning and temperature profile controls are shown in Figure 4.

Figure 4: Kiln burning and temperature profile controls. The primary objective in lime kiln optimization is to control the temperature profile from the inlet to the outlet.

There are some process limitations and disturbances, which are also addressed by the optimization controls. Using fuzzy logic, the kiln rotation speed and temperature profile targets can be changed to introduce temperature gradients, which can loosen ring buildup. Also, if excessive dusting is a problem, the kiln rotation speed can be changed. NOx generation can be minimized by proper primary to secondary air ratio.

The controls are implemented in the DCS system using function block software. These are in common use and easy to maintain and tune.

SPECIFIC ENERGY CONSUMPTION

The energy consumption in the process is not specifically controlled. One level is set to produce the desired lime quality lime. Variation in lime mud moisture is the most significant variable which can affect energy consumption, in some cases by 5 to 10%. The energy savings are realized by adapting all of the process variables in an optimum way to achieve even lime quality and while minimizing temperature targets. Any variable out of its optimum range can disrupt the whole process. The DNAlime system includes a specific energy monitoring feature which keeps track of energy use per ton of lime.

Monitoring and optimization of the energy usage of the pulp mill can minimize the energy costs more accurately in the lime kiln operation. This can be linked to the overall plant energy managements system.

Usually the lime kiln is the only department using fuel purchased outside the mill. So it is a direct savings in costs.

KRAFT MILL RESULTS

Results in kraft pulp mills attest to the stabilizing effects of lime kiln optimizing controls. At Stora Enso's Veitsiluto mill in Kemi, Finland lime, quality stability, as measured by residual calcium carbonate, has been improved by 80%. Also, the specific energy consumption has been reduced by 4%. That stability has a carry-through benefit in the causticizing plant where the degree of causticizing is now much more stable and the capacity of the plant has been increased by 19%. Figure 5 shows the results from this mill.

The most recent DNAlime optimization controls have been installed in the Smurfit Nettingsdorfer kraft pulp mill near Linz, Austria. The 240,000 tpy integrated batch–process pulp mill makes 100 Kappa number unbleached pulp for the onsite linerboard machine. The metsoDNA- based control was added to an existing Metso Automation Damatic XD system. The lime kiln was delivered by Ahlstrom. It includes one of the first lime mud drying (LMD) systems.

One of the main reasons for the investment was to achieve energy savings required by the European Union in 1995, for compliance to the Kyoto accord limiting carbon dioxide generation . Also the mill planned to improve the lime quality by achieving better stability in the process. This would alleviate capacity-limiting problems in the causticizing process.

The mill has been satisfied with the stabilizing effects of the new lime kiln controls. The mill reports that operators never realized before the outcome of a change until 8 or 12 hours later . They could never optimize that manually. Now the operators have a good trust in the controls and it manages long term changes well. It is easier for the operators to make changes with no upsets in production. It keeps the lime quality at the same level.

Rudolf Bito, production manager, says the controls have provided improved stability. "The operation of the kiln is much easier. With a delay of several hours it's complicated for the operators to control but the control system handles it well." He also appreciates the system's ability to calculate and report oxygen levels in the flue gas. He says, "The system is clever enough to calculate the correct oxygen level and select the right one, measured or calculated, to use for control." This has helped out when the oxygen sensor was unreliable.

The statistics attest to the more even operation. The variation of residual carbonate has been reduced by 76.7%. According to the statistics available, the oil consumption went down by 3.3%. There is some uncertainty about that number since the previous metering system was in error. The energy savings are generally agreed to be in the range of 3 to 5%.

The results are shown in Table1.

Figure 5: At Stora Enso's Veitsiluoto mill the lime quality and causticizing process are much more stable after lime kiln optimization controls were implemented. This stability contributed to a 19% increase in causticizing plant capacity.

Table 1: The lime burning process is much more stable at the Smurfit Nettingsdorfer mill. The data show the improvement in the stability of kiln temperature, residual carbonate and specific energy.

CONCLUSIONS

Lime kiln optimization controls have been installed in a number of kraft pulp mills with good success. These controls have achieved a stability of operation that has led to more stable and lower residual carbonate levels in the lime . This has improved causticizing plant operations and alleviated bottlenecks, leading to increase production levels. Specific energy consumptions have also been lowered, thereby saving expensive fuels .

Authors contact details:

HEIKKI IMELÄINEN
Application Specialist
Metso Automation Tampere, Finland

MAURI LOUKIALA
Business Manager
Metso Automation, Tampere, Finland

URPO LAUNONEN
Senior Sales Manager
Metso Automation, Vermont, Vic, Australia