Abstract Degree thesis Solid content in Lime Mud By Andre Flodin Process operator program – Umeå University 17/6 – 2014 Pages: 21 A thesis was made on Billerudkorsnäs production unit in Karlsborg where lime kiln in the causticising department of chemical recovery plant does not meet the required production. Causticising department is the final step in the chemical recycling unit. There, the filtered green liquor is mixed with lime to recover the cooking chemicals used in the pulp factory. The cooking liquid is called white liquor. As a residue from the recycling of the white liquor you get lime mud, calcium carbonate. Calcium carbonate is recovered in the lime recycling department to reduce the need to buy lime to the factory, thus saving money. By first washing the lime mud in the lime filter and then burning it in the lime kiln. In order to increase production through the furnace better washed lime mud which means higher solid content and so little residual alkali as possible is sought. In order to increase the dry content and reduce the residual alkali in the lime mud examinations are made on one of the lime filters by means of multivariate experimental design. This is a great tool for examining whether there are interaction effects between different factors in an effective manner. The experimental design on the lime filter examines three different parameters, density of incoming lime mud flow, speed of the filter and wash water flow to the filter. Through the variation of these three variables in a calculated pattern an optimum can be obtained. There are other factors that will affect the dry matter content and residual alkali but these are not considered to affect as much. Two attempts were made, a test experiment and a main experiment. The test experiment was made to check how well the sampling and analytical methods work. Smaller version of the experimental design was used and the results are not considered to be reliable enough to be used for optimizing the process. It was however discovered some problems. The wash water flow to the filter was not stable and no compensation was made on the lime mud flow in relation to density. For the main experiment, the design CCF - Central Composite Face Centered with three center points was used, to determine the variation between samples. Before the trial started a second wash water pump was started and during the experiment compensation was made to the lime mud flow against density. With the obtained results on the center points, the variation between them made the experiment unreliable. This may be due to several reasons. When the results are not considered reliable, it is hard to make any conclusions about how the filter should be set for maximum dryness or residual alkali. But it is possible to discern that certain factors influence more than others, wash water flow and speed of the filter seems to affect the dry matter content the most.

Industrial manufacturing companies are working in a greater extent than before to optimize processes as well as reduce variation to build in quality in their products. SCA Munksund produces kraftliners with high variability for the strength measurements bursting strength and SCT-CD, together called strength potential. Variability in these strength measurements leads to either higher quality than requested by customers or to products being scrapped due to substandard quality. The variation in the strength measurements leads to unnecessary production costs according to SCA Munksund. To investigate the variation in strength measurements of a specific grammage of a kraftliner (170 g/m2), the problem-solving methodology DMAIC in Six Sigma was used. DMAIC can be used to translate a complex production problem into a more manageable statistical problem. Design of Experiments (DoE) can be used together with DMAIC to identify casual relationships and can also be applied to increase process knowledge. DoE is a powerful tool when studying how effects of multiple process parameters affect quality parameters in a process. In this master thesis, DMAIC was applied as a method to illustrate how a complex production problem can be tackled. The purpose of this master thesis work was to exemplify how statistical tools and methods can be applied in a manufacturing industry to achieve increased process knowledge. The purpose was applied practically at SCA Munksund kraftliner production site to create increased understanding of the subprocesses where the strength measurements bursting strength and SCT-CD can be affected. DMAIC was initially used to define the problem and DoE was used to generate models that could explain the strength potential. To help answer the thesis question, historical data were analysed to identify potential experimental factors. Appropriate experimental levels were checked in a test before the experiments. Then the experimental factors for bursting strength were determined to A: Läppläge vertikal topplåda, B: Alun bas and C: Koncentration inloppslåda bas and the factors for SCT-CD to A: Läppläge vertikal topplåda, B: Utloppsdifferens top and C: Utloppsdifferens bas. The experiments used a 23 factorial design with three center points. Alun bas and Utloppsdifferens top could on a 10 percent significance level be identified as significant effects. Regression models were developed to describe which factors and factor interactions that affect the strength potential in a kraftliner. The case study results thus illustrates how statistical tools and methods can be applied in a manufacturing industry to increase process knowledge. The result show that the bursting strength and SCT-CD increased with 2,4 % and 1,3 % respectively compared to normal levels during the experiment at the same time as lower operating costs can be achieved. SCA Munksund can save 143 700 SEK per year if the recommended level of Alun bas is followed. / Industriellt tillverkande företag arbetar i större utsträckning än tidigare med optimering av processer samt för minskad variation kring uppsatta toleransgränser i syfte att bygga in rätt kvalitet i sina produkter. SCA Munksund producerar kraftliner med hög variation för styrkemåtten sprängstyrka och SCT-CD. Variationen leder till att kraftlinerrullar antingen kan säljs med bättre kvalitet än efterfrågat eller så kasseras de på grund av bristande kvalitet. I båda avseendena anser SCA Munksund att variationen för styrkemåtten leder till onödiga kostnader vid produktionen. För att utreda variationen i styrkemåtten hos en specifik ytvikt av kraftliner, 170 g/m2, användes Sex Sigmas problemlösningsmetodik DMAIC som förkortat står för Define, Measure, Analyze, Improve och Control. DMAIC används för att översätta ett komplext produktionsproblem till ett mer hanterligt statistiskt problem. Inom problemlösningsmetodiken kan försöksplanering användas för att identifiera kausala samband och kan tillämpas för ökad processkunskap. Försöksplanering är ett kraftfullt verktyg för att kunna studera hur effekterna av flera processparametrar påverkar kvalitetsparametrar i en process. Vid examensarbetet användes DMAIC för att angripa ett komplext problem. Syftet med examensarbetet var att exemplifiera hur statistiska verktyg och metoder kan tillämpas inom en tillverkande industri för att nå ökad processkännedom. Syftet applicerades praktiskt vid SCA Munksunds kraftlinertillverkning för att skapa ökad förståelse inom de delprocesser där styrkepotentialen kan påverkas. DMAIC användes för att initialt kunna definiera problemet och försöksplanering användes för att skapa regressionsmodeller som kan förklara styrkepotentialen. För att uppnå syftet analyserades historisk data för att identifiera potentiella försöksfaktorer. Lämpliga försöksnivåer kontrollerades vid ett test innan experimenten. Därefter kunde försöksfaktorerna fastställas för sprängstyrkan till A: Läppläge vertikal topplåda, B: Alun bas och C: Koncentration inloppslåda bas och faktorerna för SCT-CD till A: Läppläge vertikal topplåda, B: Utloppsdifferens topp och C: Utloppsdifferens bas. Experimentet genomfördes enligt en 23 försöksplan med tre centrumpunkter. Alun bas och Utloppsdifferens toppkunde med 10 % signifikansnivå identifieras som signifikanta effekter. Anpassade regressionsmodeller skapades i syfte att beskriva vilka faktorer och samspel mellan faktorer som påverkar styrkepotentialen i kraftlinern. Fallstudien illustrerar därmed hur statistiska verktyg och metoder kan tillämpas inom en tillverkande industri för att nå ökad processkännedom. Jämfört mot normala nivåer visar resultaten från de experiment som genomfördes i fallstudien att sprängstyrkan ökade med 2,4 % och SCT-CD med 1,3 % samtidigt som lägre driftskostnader kan uppnås. Efterföljs den rekommenderade nivån för alun i kraftlinerns basskikt leder det till en besparing på 143 700 SEK per år. Besparingen av alun är endast beräknad på den studerade kraftlinern med ytvikt 170 g/m2, vilket innebär att ytterligare besparingar bör vara möjliga för resterande 18 ytvikter.