Total Site Analysis (TSA) Stenungsund

Research Project Report

ROMAN HACKL AND SIMON HARVEY

Department of Energy and Environment Division of Heat and Power Technology CHALMERS UNIVERSITY OF TECHNOLOGY

EVA ANDERSSON
CIT Industriell Energianalys AB

PROJECT REPORT

Total Site Analysis (TSA) Stenungsund

ROMAN HACKL SIMON HARVEY EVA ANDERSSON

PROJECT LEADER: Simon Harvey

Department of Energy and Environment Division of Heat and Power Technology CHALMERS UNIVERSITY OF TECHNOLOGY Gteborg, Sweden 2010

Total Site Analysis (TSA) Stenungsund ROMAN HACKL SIMON HARVEY EVA ANDERSSON

Department of Energy and Environment Division of Heat and Power Technology Chalmers University of Technology SE-412 96 Gteborg Sweden Telephone: + 46 (0)31-772 1000

Cover: Satellite view of the chemical cluster in Stenungsund, Sweden

Total Site Analysis (TSA) Stenungsund

Summary
This project was carried out in cooperation between the Division of Heat and Power Technology at Chalmers University of Technology, CIT Industriell Energianalys AB, AGA Gas AB, Akzo Nobel Sverige AB, Borealis AB, INEOS Sverige AB and Perstorp Oxo AB. A Total Site Analysis (TSA) was performed in this study which can be used as a basis for future implementations of energy system integration at the chemical cluster in Stenungsund. At first stream data (Tstart, Ttarget, Q) and data on overall utility consumption of all the processes in the cluster was collected. The analysis is based on data collected on process streams heated or cooled with utility exceeding a heat load of 300 kW. Additionally steam from by-product incineration which cannot be utilised in another way is considered as process heat. With this data the current energy system was analysed by determining steam excess and deficit at each steam level and company. After that, the data was represented in curves, the so called total site profiles (TSP) and the total site composite curves (TSC). The curves were used to determine the site pinch (the limiting factor for further integration) and to identify measures to increase heat recovery. The measures found by TSA were assessed qualitatively with respect to feasibility to determine the most attractive measures. Finally the site wide potential for cogeneration and measures for reduction of external cooling demand below ambient temperature was analysed. Main findings are presented in the following: From the stream data collected is can be seen that the total demand of hot and cold utility of the cluster is 442 MW and 953 MW respectively. By-products, which have to be incinerated on-site provide 40 MW of steam. To cover the external heat demand additional 122 MW of heat is supplied by steam/hot oil from boilers or directly by flue gas from added fuels purchased or available on site. The TSP and TSC curves show a site pinch at the 2 bar(g) steam system (132 C). The site pinch limits the potential for heat integration. To increase energy savings by heat integration it is necessary to change the position of the site pinch. It was shown that theoretically by introducing a site-wide hot water circuit, increased recovery of 2 bar(g) steam and adjustment steam levels in several heat exchangers the pinch point can be moved so that hot utility savings of 122 MW plus excess of 7 MW steam at 85 bar(g) can be realised. Only introducing a hot water circuit can save 51 MW of steam from added fuels, which corresponds to estimated savings of 122 MSEK/year. It is possible to replace more steam by hot water, but the demand for 2 bar(g) steam is limited. Therefore a demand for low pressure steam must be created by adjusting steam levels in order to utilise more waste heat in a hot water circuit. The present delivery of heat to the district heating system is not affected by a site wide hot water circuit. I

There is potential for increased recovery of 33 MW of 2 bar(g) steam from process heat. This would replace the production of the same amount of steam in the boilers, worth 79 MSEK/year. A qualitative assessment on the implementation of a hot water circuit shows estimated steam savings of 55.2 MW (132 MSEK/year) with moderate changes (83.5 MW including more complex changes, 200 MSEK/year). Technically the introduction of a hot water circuit includes hot water pipes between several plants, as most of the consumers of heat are situated at the cracker site and at Perstorp but the sources are spread out across the cluster. Also piping is necessary to transfer the 2 bar(g) steam replaced by hot water to other plants with steam deficit. The practical potential for increased 2 bar(g) steam recovery is estimated to 4.2 MW (10 MSEK/year) with moderate changes and 26.6 MW including more complex changes (64 MSEK/year). Increased 2 bar(g) recovery implies the construction of steam pipes from Borealis to Perstorp and INEOS, as most of the potential steam sources are located at Borealis but Perstorp and INEOS have a demand for 2 bar(g) steam The theoretical cogeneration potential in the cluster is 19 MWel in addition to the 10 MWel generated today (additional revenue is 40 MSEK/year) assuming that steam demand at all pressure levels remains the same but the steam systems are connected with each other. A practical option to increase cogeneration with the existing equipment is to supply steam below 8.8 bar(g) produced at Borealis to INEOS, Akzo and Perstorp. This would result in additional 8.6 MWel by cogeneration in Borealis turbo-alternator (estimated revenue: 18 MSEK/year). Some process streams below ambient temperature are heated with steam. It has been shown that 6.5 MW steam is used for heating stream well below ambient temperature. This steam can be saved and the cooling energy can be recovered. This decreases the energy usage in the cooling system and also saves heating steam. Savings up to 48 MSEK/year were estimated. It has been shown that by site wide collaboration it is possible to increase heat recovery, cogeneration and utilisation of waste heat. The results from this study are the bases to identify concrete projects which contribute to cost and CO2 emissions savings. The study also shows the advantages of TSA in order to find solutions for process integration by the utility system on a site wide level. Keywords: Total Site Analysis, Pinch Technology, Chemical Clusters, Area wide process integration, Utility System, Cogeneration

II

Total Site Analysis (TSA) Stenungsund

Sammanfattning
Projektet r ett samarbete mellan Vrmeteknik och maskinlra p Chalmers, CIT Industriell Energianalys AB , AGA Gas AB, Akzo Nobel Sverige AB, Borealis AB, INEOS Sverige AB och Perstorp Oxo AB. TSA-analysen (Total Site Analysis) anvnds som en grund fr att identifiera tgrder fr att integrera energisystemen vid det kemiska klustret i Stenungsund med syfte att effektivisera energianvndningen. Analysen brjade med insamling av data fr processtrmmar (Tstart, Ttarget, Q) och data fr totala nganvndningen fr alla processer i klustret. Analysen r baserad p data som samlats in fr de processtrmmar som vrms eller kyls med vrmande och kylande media och har en vrmeverfring som r strre n 300 kW. nga som produceras med biprodukter som mste eldas p anlggningen rknas som processvrme. Det nuvarande energisystemet analyserades fr att bestmma ngverskott och ngunderskott vid de olika anlggningarna. Efter det anvndes de insamlade vrdena fr att konstruera kurvor, TSA (Total Site Profiles) och TSC (Total Site Composite curve). Kurvorna anvndes fr att bestmma pinchen (den punkt som hindrar integration) och fr att identifiera tgrder som gr det mjligt att ka integrationen. En bedmning av de tgrder som framkom i TSA-analysen gjordes fr att avgra hur de r mjliga att genomfra i praktiken. Till slut gjordes en bedmning av potentialen fr tillverkning av mottrycksel och en analys av tgrder fr att minska det externa kylbehovet fr strmmar under omgivningens temperatur. Sammanfattning av resultatet: De insamlade vrdena visar att det totala behovet av vrme och kyla fr hela Stenungsundsindustrin r 442 MW respektive 953 MW. De biprodukter som mste frbrnnas lokalt ger ett tillskott p 40 MW nga. Fr att tcka det externa vrmebehovet s tillfrs 122 MW i form av nga/hetolja frn pannor eller direkt som vrme frn avgaser vid direkt frbrnning av brnsle som finns att tillg i Stenungsund eller brnsle som kps utifrn. Frn TSP- och TSC-kurvorna kan man avlsa att pinchen fr siten r vid 2 bar(g)nga (132oC). Pinchen begrnsar mjligheterna fr att integrera energisystemen. Fr att minska energianvndningen genom integration r det ndvndigt att frflytta pinch-punkten. Genom att introducera en hetvattenkrets, ka genereringen av 2 bar (g)-nga och anpassa ngniverna i ett flertal vrmevxlare s kan pinchen flyttas s att den teoretiska besparingen av vrmemedia blir 122 MW plus att det blir ett verskott av 7 MW nga vid 85 bar(g). Om man bara installerar ett hetvattensystem s kan man spara 51 MW nga, vilket motsvarar en berknad besparing p 122 MSEK per r. Det r mjligt att erstta mer nga med hetvatten, men det resulterar i ett verskott av 2 bar(g)-nga. Drfr mste behovet fr lgtrycksnga kas genom att anpassa ngniver fr andra frbrukare om III

mer vrme ska kunna nyttjas i ett hetvattenssytem. Dagens leverans av fjrrvrme pverkas inte av att ett hetvattensystem infrs. Det finns en potential fr att tillverka ytterligare 33 MW 2 bar(g)-nga vid kylning av processen. Detta skulle kunna erstta nga som produceras i ngpannor och kan vrderas till 79 MSEK per r. Vrdering av den praktiska genomfrbarheten av ett hetvattensystem visar att man med inte allt fr stora ombyggnader kan erstta 55.2 MW nga (132 MSEK per r) och med strre ombyggnader 83.5 MW nga (200 MSEK). Tekniskt innebr introduktionen av ett hetvattensystem att hetvattenledningar byggs mellan mnga av anlggningarna eftersom efterfrgan p hetvatten r strst p Borealis krackeranlggning och Perstorp medan kllorna till hetvattensystemet finns p alla anlggningar. Det innebr ocks att man behver bygga ngledningar fr 2 bar(g)nga fr att kunna transportera den nga som sparas till de anlggningar som har ett behov av lgtrycksnga. Den praktiska genomfrbarheten nr det gller mer generering av 2 bar(g) nga frn processvrme r 4.2 MW (10 MSEK per r) fr de mindre omfattande tgrderna och 26.6 MW (64 MSEK per r) om man inkluderar mer omfattande tgrder. Om mer 2 bar(g) nga generas i processen s innebr det att det krvs ngledningar frn Borealis till Perstorp och INEOS, eftersom den mesta ngan kan genereras p Borealis medan Perstorp och INEOS har behov av 2 bar(g) nga. Den teoretiska potentialen fr att producea mottrycksel vid Stenungsundsindustrierna r 19 MWel i tillgg till de 10 MWel som genereras idag (vilket skulle innebra 40 MSEK per r) frutsatt att behovet av nga p olika tryckniver r detsamma som idag och att ngsystemen r integrerade. Ett praktiskt alternativ fr att ka produktionen av mottryck r att hela ngbehovet under 8.8 bar(g), som finns hos INEOS, Akzo och Perstorp, frsrjs med nga frn Borealis som kan producera mottrycksel i turbo-alternatorn. Det skulle resultera i ytterligare 8.6 MWel (18 MSEK per r). Ngra processtrmmar som opererar under omgivningens temperatur vrms med nga. Det r 6.5 MW nga som anvnds fr att vrma strmmar som r lngt under omgivningens temperatur. Denna nga kan sparas och dessutom kan kylan anvndas fr att tcka kylbehov i andra processtrmmar. Detta innebr besparingar i bde energi till kylsystemet och nga. Besparingarna uppskattas till 48 MSEK per r. Studien har visat att samarbete mellan anlggningar kan ka vrmetervinning, produktion av mottrycksel och anvndning av tervunnen vrme. Resultatet frn detta arbete kan ses som grunden fr att identifiera konkreta projekt som bidrar till minskade kostnader och CO2-utslpp. Studien visar frdelarna med TSA-analys nr det gller att hitta lsningar fr att utveckla gemensamma system fr vrmande och kylande medier ver anlggningsgrnser. Nyckelord: Total Site Analysis, Pinchteknik, Kemikluster, Processintegration, Servicesystem, Mottryckskraft

IV

Contents
SUMMARY SAMMANFATTNING CONTENTS PREFACE NOTATIONS 1 2 3 INTRODUCTION OBJECTIVE PROCESSES AND UTILITY SYSTEMS DESCRIPTIONS 3.1 AGA Gas 3.1.1 Products and processes Utility System 3.1.2 3.2 Akzo Nobel Products and processes 3.2.1 3.2.2 Utility system 3.3 Borealis 3.3.1 Products and processes Utility system 3.3.2 3.4 INEOS Products and Processes 3.4.1 3.4.2 Utility system 3.5 Perstorp Oxo 3.5.1 Products and processes 3.5.2 Utility system 4 METHODOLOGY 4.1 Pinch Analysis 4.1.1 General 4.1.2 Composite curves Grand Composite Curve 4.1.3 I III 1 3 4 5 6 7 7 7 8 8 8 8 9 9 10 11 11 11 12 12 12 13 13 13 14 14

4.2 Total Site Analysis (TSA) 15 4.2.1 Choosing level of detail of the total site analysis 15 4.2.2 Graphical representation of the current steam systems 16 4.2.3 Representing the process streams with site source sink profiles (SSSP) 18 4.2.4 Analysing utility use with total site profiles (TSP) 19 4.2.5 Determine heat demand with total site composites (TSC) 19 4.2.6 Identifying improvements to the utility systems 20 4.2.7 Assessment of potential energy savings accounting for practical constraints 20 1

4.2.8 Quantifying impact of energy efficiency measures on cogeneration potential. 21 5 RESULTS 5.1 The current utility system 5.1.1 Steam production 5.1.2 Steam use 5.1.3 Handling of other steam users 22 22 22 23 23

5.2 Identifying improvements to the utility systems 24 5.2.1 Increased energy efficiency by integrating the existing steam systems 24 5.2.2 Using total site profiles to improve energy efficiency 27 5.3 Qualitative evaluation of potential solutions 5.3.1 Hot water circuit 5.3.2 Increased steam recovery 5.3.3 Further changes to the utility system to increase savings 36 36 38 39

5.4 Cogeneration potential 39 5.4.1 Theoretical cogeneration potential with the current utility levels and 39 current steam demand and excess in an integrated steam system 5.4.2 All steam is produced at Borealis and distributed to plants with a 39 deficit of steam below 8.8 bar(g). 5.4.3 Cogeneration potential with a totally integrated steam system and a new, common boiler 40 5.4.4 Cogeneration at Perstorp 41 5.5 6 7 8 9 Heat exchangers operating below ambient temperature 41 43 45 46 1 1 1 3 4

CONCLUSIONS FUTURE WORK REFERENCES APPENDIX 9.1 9.2 9.3 9.4 Appendix 1 Appendix 2 Appendix 3 Appendix 4

Preface
This project was carried out in cooperation between the Division of Heat and Power Technology at Chalmers University of Technology, CIT Industriell Energianalys AB, AGA Gas AB, Akzo Nobel Sverige AB, Borealis AB, INEOS Sverige AB and Perstorp Oxo AB. Financing of the project was provided by the participating companies (60%) including their working contribution and by the Swedish Energy Agencys Process Integration programme (40%). The project was conducted between October 15th and May 15th 2010 with Simon Harvey (Professor of Industrial Energy Systems at Chalmers University of Technology) as project leader. The study includes a total site analysis of the chemical cluster in Stenungsund and further investigations on increased energy integration, common utility systems for the participating companies and the cogeneration potential on site. The following people participated in the study. Jerker Arvidsson at AGA Gas, David Ekeroth and Erik Falkeman at Akzo Nobel, Reine Spetz at Borealis, Kent Olsson, Peter Andersson and Heln Axelsson at INEOS, Milos Pollard and Lars Lind at Perstorp, Simon Harvey at the Department of Energy and Environment, Chalmers and Per-ke Franck at CIT Industriell Energianalys. Wed like to thank all the participants for their support and the participating companies and the Swedish Energy Agency for the financing of this project.

Notations
Abbreviations: Ar BA-10 CC CCS DEA EDC EMU EO ETBE GCC H2 HCl HDPE HP HPPE LDPE LLDPE LP LPPE MEA MP NaOH PE PVC RME SCN SSSP STF TSA TSC TSP VCM VHP Symbols: Q Qcooling,min Qcooling,total Qheating,min Qheating,total Qrec Tstart Ttarget Tmin Heat load Minimum cooling demand Overall cooling demand Minimum heating demand Overall heating demand Heat recovered Stream starting temperature Steam target temperature Minimum temperature difference 4 Argon Berolamine-10 Composite Curves Carbon Capture and Storage Diethanolamine Ethane dichloride Emulgol Ethylene oxide Ethyl Tertiary Butyl Ether Grand Composite Curve Hydrogen Hydrogen chloride High density polyethylene High pressure High pressure polyethylene Low density polyethylene Linear low density polyethylene Low pressure Low pressure polyethylene Monoethanolamine Medium pressure Sodium hydroxide Polyethylene Polyvinylchloride Rapeseed oil methyl ester Steam cracked naphtha Site Source Sink Profiles Speciality surfactants Total Site Analysis Total Site Composites Total Site Profiles Vinyl chloride Very high pressure

1 Introduction
The chemical cluster in Stenungsund is Swedens largest agglomeration of its kind. The main companies involved are AGA Gas AB, Akzo Nobel Sverige AB, Borealis AB, INEOS Sverige AB and Perstorp Oxo AB. The heart of the cluster is a steam cracker plant run by Borealis. Figure 1 shows the material flows between the different companies and plants.
Ethylene Ethylene Import terminal Oxygen Air Nitrogen Fuel gas
Hydro Polymers

Fuel gas
Polyethylene plant

Polyethylene

Polyvinyl chloride

Hydrogen gas Fuel gas Fuel gas Naphtha Ethane Propane Butane
Cracker plant

Ethylene

Amines Tensides

Electricity Water

Hydrogen gas
Perstorp Oxo AB

Chemical intermediates

Propylene
Natural gas

Figure 1 Major material and energy flows in the Stenungsund chemical cluster (Borealis AB 2009)

As can be seen in Figure 1 the companies already interact strongly with each other in terms of material exchange. This cooperation shall be complemented by energy integration throughout the chemical cluster in Stenungsund. Chalmers has undertaken a number of studies conducted as M.Sc. thesis projects in the area of energy integration in collaboration with some of the participating companies. Such studies have quantified the potential for increased energy efficiency by process integration for some of the participating companies production sites. Since these studies were mostly limited to single companies, the potential for energy saving by increased intercompany cooperation in terms of energy integration has not previously been addressed. Site-wide process integration studies within industrial clusters often show large potential for energy savings, on average 20-25% compared to the current energy consumption of the total site (Linnhoff March 2000). Such studies provide the opportunity to analyze integration of additional energy processes making use of the site infrastructure, thereby contributing to an increased overall efficiency of the site. In order to achieve these goals, total site analysis (TSA) can be used as a tool to analyse the energy situation in an industrial cluster. TSA produces targets for the amount of utility used and produced through energy recovery by the different processes. The method enables investigation of opportunities to deliver waste heat from one process to another using a common utility system. It also enables redesign of the utility system for increased efficiency, e.g. adjusting steam levels to fit the site requirements. 5

2 Objective
The objective of this study is to perform a total site analysis based on pinch analysis in order to quantify the site-wide potential for increased energy efficiency at the chemical cluster in Stenungsund. The method is based on identifying process streams that are heated or cooled with utility in order to find possibilities for increased integration between utility systems of the plants. The study also aims to suggest practical ways to achieve a more integrated utility system in order to increase energy savings, based on the results of the total site analysis. Since the current utility system includes steam turbine cogeneration, an important goal of the project is to quantify the impact of possible changes to the total site utility system on the cogeneration potential. The study also aims at increasing knowledge about the chemical cluster, which can be used as basis for further studies, such as integration of energy-intensive climatefriendly processes such as advanced biorefinery concepts or Carbon Capture and Storage (CCS) for the cluster.

3 Processes and utility systems descriptions

At the first site visit at each company we received a brief presentation of the processes and the utility systems. Below is a short presentation of the included companies. We also defined the process stream data that had to be provided by the companies in order to enable us to perform the total site analysis. Before initiating the collection of data we decided to limit the number of heat exchangers by only including heat exchangers with a heat load above 300 kW. This was done in order to handle the project within the given timeframe. To be able to construct the composite curves consisting of process streams and the utility curves the following data was requested from the companies for each heat exchanger with a heat load > 300 kW : Name or identification of heat exchanger Start temperature of process stream Exit temperature of process stream Heat exchanger heat load (or data necessary to calculate the heat load) Utility used for heating or cooling

Data was requested to be representative for a normal operation situation. After analysing the collected data we had another meeting at the sites to evaluate the possibility to implement the modifications identified as interesting in order to increase heat recovery from the process and reduce the added fuel demand.

3.1 AGA Gas

3.1.1 Products and processes
AGA Gas is a manufacturer of industrial gases and operates a cryogenic air separation plant in Stenungsund. The plant produces oxygen (O2), nitrogen (N2), carbon dioxide (CO2) and argon (Ar) for the local companies and for sale on the market. In a first stage the air is compressed and then purified by a molecular sieve to remove water and CO2. Then the incoming feed air is heat exchanged with the outgoing product gases and waste streams in order to recover refrigeration from the leaving streams. After heat exchanging the air has a temperature of app. -175C before it enters the cryogenic distillation column. Here N2 leaves as top product while O2 leaves the column at the bottom. To produce high purity O2, an Ar rich side stream is removed from the low pressure column. The Ar rich stream can be vented or purified by oxidising hydrogen to water and removing this in a molecular sieve drier. The product streams are routed back to the front end heat exchanger to cool the incoming air and are thereby warmed to near ambient temperature. The liquid products can also be sent to storage tanks, from where liquefied N2 and O2 can be loaded on trucks or stored as back-up in case of varying demand at other plants in Stenungsund (UIGI 2009). Industrial CO2 is produced from a CO2-rich stream at Akzo Nobels plant which is purified by AGA.

3.1.2

Utility System

Aga gas has a very low steam consumption and imports 28 bar(g) steam from Akzo Nobel. The steam is reduced and used at a low pressure. Cooling water and an ammonia system is used as external cooling utility. The largest energy consumption is electricity with app. 23 MW for compression of air and product gases.

3.2 Akzo Nobel

3.2.1 Products and processes
Akzo Nobels site in Stenungsund consists of three plants. The heart of the site is the ethylene-oxide (EO) plant. EO is used as a raw material in the production of among others ethyloxylates, cellulose derivatives, ethanol- and ethylene amines. It is produced by partial oxidation of ethylene. Thereby EO, carbon dioxide (CO2) and water are formed. The reaction is strongly exothermic. Steam is produced when cooling the reactor. After reaction the EO is dried, purified and sent to storage. Part of the EO is further reacted with water in a glycol reactor to form mono-ethylene glycol as the main product, which is used e.g. as in antifreeze or solvents. The CO2 is converted into carbonic acid or directly sold to customers. In the surfactants plant surface agents are produced, which are used in e.g. disinfectants, textile softeners and detergents. Two units, the emulgol (EMU) and speciality surfactants (STF) unit, manufacture over 300 different end-products. The EMU unit converts EO (or propylene oxide) together with fatty amides, amines, alcohols and acids into end-products. The STF unit applies a large variety of reactions generating different groups of surfactants. The amine plant consists of two main processes. In one process ethanol amines are produced by a reaction between EO and ammonia. After the reaction the products are separated into diethanolamine (DEA), monoethanolamine (MEA) and Berolamine-10 (BA-10). MEA is further converted with ammonia and hydrogen in a catalytic reaction under high pressure and high temperature into ethylene amines in the ethylene amine process. A number of products are leaving the reactor and are purified in several separation units (Akzo Nobel 2004).

3.2.2

Utility system

The steam pressure levels at Akzo are 40, 28, 20 and 6 bar(g). Steam at 40 bar(g) is generated in a boiler fired with fuel gas, imported from Borealis and in a waste incinerator where by- products are fired. There is another boiler that delivers steam to the 28 bar(g) system. 20 bar(g) steam is generated from heat recovery from cooling of exothermic reactors. Akzo recovers heat in an internal hot water system that operates between 65-85C. Sea water is used to cool the processes and cooling compressors are used to cool ethylenoxid storage below ambient temperature. In 2011 the cooling compressor system will be replaced by internal cooling energy from Akzos ammonia terminal where ammonia is vaporised.

3.3 Borealis
3.3.1 Products and processes
Borealis is a supplier of plastic material for wires and cables, pipes, automotive and advanced packaging. In Stenungsund two Borealis sites are in operation. A cracker plant produces olefins and three polyethylene (PE) plants produce different qualities of PE used primarily for pipes, wires and cables. The cracker plant provides the feedstock to the Borealis PE plant, and to other companies present in the Stenungsund chemical cluster. The nine cracking furnaces in the cracker plant convert a feedstock of naphtha, ethane, propane and others into unsaturated hydrocarbons, hydrogen, fuel-gas, cracked gasoline and heavier products. The different products from the cracking processes are separated downstream mainly by distillation, cooling, compression and further reactions before being stored in tanks or directly delivered to customers. ETBE is produced from a share of butylene/butadiene leaving the cracking furnaces in another plant. Table 1 shows the inputs and outputs from the cracker plant for the year 2008.
Table 1 Feedstock consumption and production of the cracker in 2008 (Borealis AB 2009)

Feedstock Naphtha Ethane Propane Butane Ethanol Off-gas

[ktonnes/year] 298 321 286 268 7 3

Product Ethylene Propylene Other (incl. fuel gas) ETBE

[ktonnes/year] 565 174 424 16

The PE area consists of three low pressure units (LPPE) based on a catalytic polymerization in gas phase and loop reactors to manufacture High Density PE (HDPE) and Linear Low Density PE (LLDPE) and a high pressure process (HPPE) producing low density PE (LDPE). In the low pressure units the incoming raw materials are cleaned before entering a circulating gas reactor. The reaction is exothermic. Circulating gas is cooled in water or air coolers respectively. One of the units works with the PE3 Borstar process which produces bimodal 1 HDPE (Spetz et al. 2006). The PE3 Borstar plant additionally contains two loop reactors (prepolymerisation and main loop) and gas recovery facilities where unreacted raw material are purified and fed back to the process. The current high pressure process will soon be phased out and replaced by another high pressure plant which will start operation in 2010. In the 5-stage primary and 2stage hyper compressor the ethylene together with moderators is compressed up to 3200 bar. Recovered ethylene from the process is recompressed in the first two stages of the primary compressor and the mixed with fresh ethylene and additives. Process gas is pre-heated to 150-180C. Polymerisation takes place at max. 350C. After the reactor the product is cooled and separated from unreacted ethylene in a high and low
1

A multimodal polymer includes more than one molecular weight fraction. Properties can be better adjusted to specific applications (Whiteley 2000)

pressure separator and the degassing silos. High pressure ethylene is cooled and separated from waxes before entering again the hyper-compressor. The low pressure ethylene is separated from waxes, cooled and fed to the primary compressor. The separated PE goes to the extruder line where it is further processed into pellets and sent to storage (Meyers 2004, chap.14). Table 2 shows the inputs and outputs from the PE units for the year 2008.
Table 2 Feedstock consumption and PE production in the PE units in 2008 (Borealis AB 2009)

Feedstock Ethylene Propylene Co monomer Additives Master batch

[ktonnes/year] 444 0.132 8.57 4.3 17.3

PE production per unit High Pressure PE (LDPE), to be phased out in 2010 Low Pressure PE (HDPE) PE3 Borstar, (HDPE) New HPPE plant (LDPE), future production

[ktonnes/year] 150 93 206 350 2

3.3.2

Utility system

Fuel gas is used to heat the cracker furnaces, the steam boilers, the pilot gas system to the flares and the hot oil furnace. The fuel gas used consists of by-products from the cracking process, the ETBE production and if necessary natural gas, ethane and propane. The steam pressure levels at the cracker plant are 85, 8.8, 2.7 and 1.8 bar(g). Steam at 85 bar(g) is generated by cooling the cracking products and in steam boilers. By expanding 85 bar(g) steam to 8.8 bar(g) several turbines are driven. The 8.8 bar(g) steam is mainly used as direct steam to the cracking furnaces, for heating purposes and as driving steam for turbines in the cracker plant. The back pressure of the turbines is 1.8 bar(g). This steam is used to supply heat to the cracker and in the deaerator of the boiler feed water system. In a 2.7 bar(g) steam system heat is recovered from the quench oil coolers after the cracking furnaces and delivered to several reboilers. Hot oil is heated in a furnace to 277C and supplied to the SCN unit (Hedstrm & Johansson 2008). The steam levels at the PE plant are 40, 11, 4 and 3 bar(g). 40 bar(g) steam imported from the cracker and produced in a boiler by incinerating by-products. 11 bar(g) and 4 bar(g) steam is recovered from the new HPPE and used at the PE plant. Excess steam can be exported as 4 bar(g) steam to the cracker plant. 3 bar(g) steam covers most of the heating demand of the LPPE units. Sea water and air are used for external cooling at the cracker plant. As the products from the cracker process have to be cooled significantly below atmospheric temperature two refrigeration systems (ethylene and propylene) are operated each generating three different temperature levels. The propylene system operates at 9C, -21C and -40C. The ethylene system used for even further cooing works at -62C, -84C and -100C (Hedstrm & Johansson 2008). Cold utilities at the PE plant are fresh water from cooling towers and air.

Annual capacity

10

3.4 INEOS
3.4.1 Products and Processes
INEOS ChlorVinyls in Stenungsund produces chlorine, sodium hydroxide (NaOH), hydrochloric acid (HCl), hydrogen, ethane dichloride (EDC), vinyl chloride (VCM) and polyvinylchloride (PVC) in three different plants. Table 3 shows the production of the different products in 2008.
Table 3 Production of INEOS products in Stenungsund 2008 (Josefsson 2009)

Product Chlorine NaOH HCl Hydrogen (sold) VCM EDC S-PVC (hard products) P-PVC (soft products)

[tonnes/year] 112 363 125 832 24 602 295 130 933 37 365 130 565 68 916

A chlorine plant converts sodium chlorine into chlorine, hydrogen and sodium hydroxide in an electrolytic amalgam-process. NaOH is mostly used in paper pulp manufacturing. Most of the hydrogen is combusted to produce steam, but some is sold to other companies. The chlorine is used as feed together with ethylene in a vinyl chloride plant to produce EDC and VCM. EDC is produced in two ways. One part comes from the reaction of chlorine and ethylene to EDC. Another process uses ethylene and HCl as feed. After both processes EDC is separated and sent to a cracking furnace where a thermal conversion to VCM takes place. After cracking the VCM is purified and sent to storage. VCM is the monomer of PVC. The polymerisation takes place in batch reactors. The reactors are charged and then heated to 45-75C (Allsopp W. & Vianello 2009). After the polymerisation (exothermic) started the vessels are cooled to maintain a constant temperature of 40-60C. In the next step the solid PVC is separated from unreacted VCM, which is reused in one of the reactors. PVC is then dried by different processes and sent to storage (Josefsson 2009).

3.4.2

Utility system

The steam pressure levels at INEOS are 28, 20, 10, 6 and 1 bar(g). 28 bar(g) steam is generated in a boiler that can be fuelled with gas or oil. A boiler at 20 bar(g) is fired with by-products from the process. Steam at 10 bar(g) and 1 bar(g) is generated with process heat. There is also a 6 bar(g) steam system. Direct firing of fuel gas is used for cracking and drying processes. Cold utility is cooling water (sea water), cooled cooling water and a propylene cooling system. The cold cooling water is operated at 4-7C and the propylene cooling system at -40C. 11

3.5 Perstorp Oxo

3.5.1 Products and processes
Perstorp Oxo AB is a manufacturer of speciality chemicals with production sites in Nol and Stenungsund. A large part of the production is iso- and n-butylaldehyde, which is an intermediate in the production of polyols applied in water based colours and coatings, security glass and softeners in vinyl products. Waste heat from the site in Stenungsund is delivered to the local district heating system which distributes heat to apartments, houses and community buildings. Synthesis gas is an important building block in processes at Perstorp Oxo. It consists of 50 % CO and 50 % H2. In a synthesis gas generator methane is partly oxidised at 1400C and 40 bars. After the reactor the synthesis gas is cooled in several steps. Major part of the heat is used to produce 40 bar(g) steam which is used to heat other processes. Some of the H2 is separated from the synthesis gas and used in other hydration processes. In two reactors aldehydes are catalytically produced from synthesis gas, H2, ethylene and propylene. In three distillation columns the products are separated after the reaction before they are sent to storage. Some of the aldehydes are sold directly while others are processed further to e.g. softeners, colours and coatings. Among other products, like octanoic acid (component in plastic films) and propionic acid (preservative agent) Perstorp manufactures rapeseed oil methyl ester (RME), which is used as vehicle fuel and glycerol an additive in food production in its RME plant (Pollard 2010).

3.5.2

Utility system

The steam pressure levels at Perstorp are 41, 20, 14, 7 and 2 bar(g). Steam at 41 bar(g) and 14 bar(g) is generated in boilers fired with fuel gas and by-products from the process. Steam to the 41 bar(g) system is also supplied from a waste heat boiler, where heat from the process is recovered. 2 bar(g) steam is generated from process waste heat. Perstorp recovers heat from the processes in a hot water system that operates between 40-120oC. The heat is supplied to other parts of the processes and up to 15 MW are delivered to the district heating system of Stenungsund. Fresh water from cooling towers and air coolers are used as cold utility.

12

4 Methodology
4.1 Pinch Analysis
4.1.1 General
Streams in industrial processes often have to be either heated or cooled in order to fulfil the process requirements. Therefore energy is added or removed respectively. This is achieved by heat exchange with hot and cold utility streams or by transferring heat between hot process streams, which have to be cooled and cold process streams which have to be heated. In order to do this in an energy-efficient way, pinch technology can be applied to design a heat exchanger network. This tool among others consists of tools and methods to quantify the minimum heating and cooling requirements of a process, define the optimal utility levels, design heat exchanger networks, estimate the cogeneration potential or heat and power integration and gives guidance for the thermal integration of equipments like e.g. distillation columns. The method is used both in grass-root and retrofit projects. The basic principle of pinch analysis is that heat withdrawn from a process or stream operating at high temperature can be used to heat another stream with a heat demand at a lower temperature. A temperature, the so called pinch-temperature is defined as the temperature above the process has an energy deficit. Below the pinch temperature the process shows a surplus of energy. From this definition the 3 golden pinch rules can be derived. Dont transfer heat through the pinch Dont withdraw heat above the pinch Dont add heat below the pinch

If these rules are broken, one commits a so-called pinch rule violation, which leads to an increased energy use compared to the minimum demand. Pinch analysis is used to determine: The minimum heating and cooling demands for the process The maximum possible internal heat recovery Temperature level of the pinch point which provides useful information for designing a cost-effective optimised heat exchanger network Guidance for selection of appropriate process hot and cold utility Guidance for the integration of energy-intensive unit operations, e.g. distillation columns Identification of possible uses of low grade waste heat (Kreith & Goswami 2007, chap.15) Guidance for integration between different processes or sub-processes Definition of the stream system Gathering stream data from appropriate sources (Tstart/Ttarget, heating/cooling loads) Definition of the minimum temperature difference for heat exchange Tmin

For a retrofit case the following steps are performed during a pinch analysis:

13

Perform pinch calculations using appropriate software (e.g. Pro-Pi3) Determine minimum hot and cold utility demand and potential for maximum internal heat exchange: Composite Curves (CC) and Grand Composite Curve (GCC) are constructed Determine the present heating and cooling demand from available process data Suggest new solutions for heat exchangers networks to eliminate pinch violations. Calculation of investment costs of suggested changes and analysis of economic performance of different options (Franck 1997)

4.1.2

Composite curves

When performing a pinch analysis one combines the temperature characteristics of all cold streams in order to construct a single composite curve. The same is done for all hot streams. Both composite curves are plotted together in one diagram, which shows the temperature versus the heat flow (T-Q-diagram). The region where the two curves overlap shows the potential heat recovery from the process. The diagram also shows the minimum heating and cooling demand (QH,min and QC,min) of the system for Tmin. This is shown by the non overlapping areas in the diagram in Figure 2. The point where the distance of the curves on the temperature axis is Tmin is the pinch point. If the two curves intersect, the cold composite curve has to be shifted to the right, to maintain Tmin. This decreases the potential internal heat exchange and leads to higher utility consumption.

Pinch point

Figure 2 Composite curves showing the Pinch point and the energy targets (Linhoff & Sahdev 2002)

4.1.3

Grand Composite Curve

Another diagram derived from the stream data and used in pinch technology is the GCC. The whole process is divided in temperature intervals. The diagram shows the heat supply and demand in each of these temperature intervals. A positive slope

Software developed by CIT Industriell Energianalys AB. Tstart, Ttarget, flowrate, cp and Q of all the streams are fed into an Excel sheet and the program calculates minimum heating/cooling demand, pinch temperature and constructs among others CC and GCC curves (Hedstrm, et al., 2008).

14

indicates a heat demand, a negative a heat surplus. This surplus can be transferred downwards the cascade to streams with heat deficit. It is amongst others used for e.g. Determining the utility levels and demands of e.g. exhaust gases, steam and cooling water Determining the potential for steam production, district heating and other energy recovery measures Integration of energy intensive unit operations, e.g. distillation columns Integration of heat pumps

Figure 3 shows an example of a typical GCC with suggestions for utility levels, internal heat exchange (pockets) and steam production (Smith 2005, p.373).

Pocket

Figure 3 Grand Composite Curve (GCC) with hot/cold utility levels and internal heat exchange

The final goal of a pinch study is to maximize the energy efficiency, reduce the costs and enhance the overall environmental performance of a single or a whole set of industrial processes. In this study pinch technology is used to identify the minimum heating and cooling demand of each plant to give an overview of the potential energy savings. Later GCC is e.g. used to determine the minimum added steam demand in a common system at the total site.

4.2 Total Site Analysis (TSA)

4.2.1 Choosing level of detail of the total site analysis
TSA is used to integrate the individual heating and cooling demands of different processes at a total site. Excess heat from one process can be used as heat source in another using a common utility system. Excess heat is transferred to a common utility (e.g. steam, hot water, hot oil) (Bagajewicz & Rodera 2001) and then passed to processes with a heat deficit by the common utility system. With the tool the amounts of hot utility generated and used by the combined individual processes, the amount of 15

heat recovery in a common hot utility system, the steam demand from the boilers and the cogeneration potential can be determined. (Simon Perry et al. 2008). The process data collected can be divided into three different groups according to the level of detail of the analysis and the availability of process data. Black box approach: Unlike in regular Pinch Analysis, where process streams are characterized by their starting temperature Tstart, target temperature Ttarget and heating/cooling loads, in the black boxes approach the process is represented by its utility demand only. E.g. if a process consumes 2000 kW of 3 bar(a) steam it is represented by this heat load at the corresponding steam temperature. Other uses of steam such as steam tracing or tank heating can be represented as black boxes. Steam consumers not included in stream data collected in this study were handled by this approach. Grey box approach: This approach considers the process-utility interface and ignores processprocess heat exchange. This means, that process streams which are heated/cooled by utilities are considered in the analysis by their starting temperature Tstart, target temperature Ttarget and heating/cooling loads. This implies that the current level of integration within the single units is not changed, but enables to identify changes in the utility system, e.g. producing hot utility instead of cooling a sufficiently hot process stream with CW, or changing utility levels to match the total site profiles. This approach is used for the analysis described in 4.2.3 to 4.2.7. White box approach or Detailed Pinch According to this method a detailed pinch analysis of the plant is carried out, including the process-process heat exchangers. In this analysis the conventional Composite Curves (CC) and the Grand Composite Curve (GCC) are constructed. Thereby the optimum hot and cold utility demand can be determined for each plant. The profile of the GCC above the pinch point represents the heat sink, below the pinch point the sources of heat are represented for each of individual process (Brown 1999).

4.2.2

Graphical representation of the current steam systems

Currently the steam systems of the different plants are not connected, it is suggested to first investigate possibilities to find ways to integrate the plants steam systems. As most of the steam headers are not connected thus site wide integration is not possible. Therefore the potential exchange of steam across the total site steam systems is investigated. Amounts of steam generated and consumed at each steam level are therefore taken from the process stream tables. Only steam generated in the processes is taken into account to determine the minimum amount of steam necessary to be produced in the boilers if optimal distribution of steam across the total site is applied. Steam produced is balanced with steam consumed at each plant and pressure level. This gives the amount of heat excess/deficit at each utility level for each plant and enables to find utility exchange possibilities at the total site. Excess and deficit are represented graphically as shown below in Figure 4. 16

Plant A 85 bar(g)/300 C Pressure level [bar(g)] Plant C40 bar(g)/250 C Plant B 20 bar(g)/215 C Plant A 20 bar(g)/215 C Plant C 10 bar(g)/185 C Plant D 8 bar(g)/180 C Plant B 2 bar(g)/133 C -4000 -2000 0 2000 4000

-6000

Steam deficit [kW]

Steam excess [kW]

Figure 4 Graphical representation of steam excess and deficit at each plant and pressure level

It is hereby suggested to use pinch analysis to determine the heating and cooling demand (see GCC in Figure 5) in the steam system, where excess of steam at one pressure level is considered as hot stream and deficit of steam as cold stream. Qheating is the demand of steam that has to be added from the boilers to cover the total steam demand. The GCC also shows the steam levels at which the deficit and excess exists (illustrated in Figure 5) and can be used as guidance for optimal distribution of steam across the total site. Excess of steam can be sent to other plants or used for future expansions of the site. 350 300 Temperature 250
20 bar(g)/215C

Qheating A
B

200 10 bar(g)/185C 150 100 50 0 0 2000 4000 6000 8000 10000 Heat load (kW)
2 bar(g)/132C Steam levels with deficit of heat Steam levels with excess of heat

Qcooling,min

Figure 5 GCC of the steam system to determine the minimum added fuel demand in the boilers

To reach this minimum steam demand a steam distribution network is developed, which enables maximum steam utilisation. Options for transferring steam between the companies and pressure levels in order to increase energy savings are developed with help of diagrams like Figure 4 and Figure 5. 17

The arrows, A and B, show how steam recovered from the process can be transferred between steam levels. Steam expansion can also be done in turbines for cogeneration. Steam added to the system from the boilers (Qheating) can be produced at a defined pressure and then expanded to the steam levels with deficits (10 an 20 bar(g) in the example). This expansion can also be performed in a steam turbine and generate electricity.

4.2.3

Representing the process streams with site source sink profiles (SSSP)

To identify more significant improvements the grey box-approach is applied, complemented by additional steam consumption not gathered in the stream data. This demand is represented in the curves as the amount of utility consumed with process temperatures corresponding to the utility temperature (black box-approach). Increased process integration will only be possible by changes in the utility system. In order to complement the study with steam demand that is not covered by the stream data gathered in this study the black box-approach is used. The plants analysed produce a considerable amount of by-products, which have to be incinerated. Thereby steam is produced. Since this steam is indirectly generated by the process the flue gases from the incineration is included in the curves as a hot stream, cooled with boiler feed water. Figure 6 shows an example of a SSSP, where the red line represents the sources of heat (streams that have to be cooled) and the blue line represents the heat sinks (streams that have to be heated). 350 300 250 Temperature [C] 200 150 100 50 0 -600000 -400000 Qcooling,total -200000 -50 0 -100 -150 Heat load [kW] Sink Profile
Figure 6 Example of a SSSP

Qheating,total

-800000

200000

400000

Source Profile

Here the total heating and cooling demand of the total site can be determined.

18

4.2.4

Analysing utility use with total site profiles (TSP)

In addition to the process source/sink profiles the utility profiles can be plotted in the SSSP diagram. This enables to analyse how heat is supplied to and discharged from the processes and makes it possible to develop improved solutions for energy recovery and cogeneration. By combining the SSSP and the site utility profiles the so called total site profiles (TSP) are obtained shown in Figure 7 to the left. The source profile, the cold utility curve, the hot utility and sink profile are shown. The utility profiles represent the amount heat supplied to or discharged from the process at the different utility temperature levels. They are developed from process stream lists by allocating the utilities used to cool/heat each process stream.

Source profile

Cold utility

Hot utility

Sink profile

Figure 7 Total Site Profiles (TSP) and Total Site Composites (TSC) (Zhu & Vaideeswaran 2000)

These profiles represent the interaction of the process heat sources and sinks with the site utilities. It also shows the total sites heating and cooling requirements and the utilities used to satisfy those requirements.

4.2.5

Determine heat demand with total site composites (TSC)

In order to find the maximum amount of heat recovery for the total site by the utility system the total site profiles are moved towards each other until the hot and the cold utility curve intersect in one point, see Figure 7 to the right. This point is the so called site pinch which limits the heat that can be recovered by the utility system (Zhu & Vaideeswaran 2000). These curves can be used to optimise the total site heat integration through the utility system by changing the utility levels to match better the process requirements and thereby increasing the overlap between the curves. The remaining minimum heating demand (Qheating) shown as very high pressure steam (VHP) in Figure 7 has to be added externally as hot utility. Qheating therefore directly relates to the fuel requirement. The cooling demand, shown as cooling water (CW) represents the amount of heat that has to be discharged from the processes to the atmosphere. Here excess heat is removed from the process by cooling media. The TSC can be used to determine the minimum necessary amount of steam produced in the boilers, potential heat recovery and cogeneration (J. Klemes et al. 1997).

19

4.2.6

Identifying improvements to the utility systems

After investigating the potential for energy savings by integrating the current utility levels, improvements by changing utility levels and introducing new ones are investigated. The profiles determined by TSA are used to investigate options to optimise the utility system. In order to increase energy recovery (save fuel) the site pinch has to be shifted. This is done by shifting utility loads as shown in Figure 8 where a new steam mains is introduced which distributes recovered steam to consumers. Changes and their effects on the overall energy situation at the total site are described. Examples for changes in the utility system are: Replacing steam by introducing a hot water circuit (Bagajewicz & Rodera 2001) Introduction of new steam mains Steam generation at higher levels Steam use at lower levels (Raissi 1994, chap.9)

Figure 8 TSC illustrating increased energy savings by modifications to the utility system, here e.g. introduction of new steam mains (Raissi 1994, chap.7)

The grey shaded boxes indicate cogeneration potential. Figure 8 shows how energy savings resulting in lower steam demand will reduce the cogeneration potential.

4.2.7

Assessment of potential energy savings accounting for practical constraints

After identification of possible changes to the utility system the practical potential for energy savings by changes in the utility system is assessed by discussions with process engineers at the respective site. The modifications options identified using the TSA analysis are qualitatively assessed using a grading system with three stages is applied: A. possible: no practical constraints, moderate investment costs are expected B. possible, with constraints: several or more complex changes are necessary, higher investment costs are expected C. impossible: process constraints make changes impossible The categories do not include any quantification of cost and the resulting practical potential has to be investigated further to find the economic potential.

20

By applying this grading system, a more practical potential for energy savings by changing the utility system is presented.

4.2.8

Quantifying impact of energy efficiency measures on cogeneration potential.

Steam is produced in boilers at a high pressure. The steam can be delivered at a lower pressure either by expanding the steam through a throttle valve or via a turbine. If the steam is throttled there is no enthalpy loss and the steam at the lower level will be superheated. Since saturated steam should be used for heat transfer processes the steam should be desuperheated. The steam is desuperheated by adding boiler feed water that will evaporate and extra steam will be generated. If the steam pressure is reduced via a turbine, electricity can be cogenerated, which is an energy efficient way to produce electricity if a heat demand is at place.
40 bar, 300 oC h= 2960 kJ/kg 10 ton/h BFW 0.7 ton/h 10.7 ton/h 10.7 ton/h 6 bar, 175 oC h= 2800 kJ/kg
Figure 9 Example of how different ways of expanding steam between two pressure levels in the utility system either can generate more steam when the steam is expanded through a valve or generate electricity if the steam instead is expanded through a turbine.

10.7 ton/h ~ 452 kW

Figure 9 shows the difference when steam is changed from one state to another along two different routes. In the first case, superheated boiler steam is throttled to utility pressure. Boiler feedwater must be mixed with the throttled steam in order to achieve desuperheating, resulting in a higher steam flow. In order to end up with the same amount of utility steam to cover the process demand, more steam has to be produced in the boiler in the second case where steam is expanded through the turbine. This extra steam is converted to electricity with high efficiency. The cogeneration potential is calculated with the following assumptions: High pressure steam properties: 85 bar(g) and 460 oC Steam from internally generated steam at 40 bar(g) is assumed to be superheated to 350 oC Steam in headers is superheated by 20 oC . When electricity is produced in a condensing turbine, the back pressure is 0.1 bar(a). Isentropic efficiency: is =0.8 Electrical and mechanical efficiency in the generator : gen=0.95

21

5 Results
This section starts with a description of the current utility system. To identify improvement to the utility sytem, the total site is analysed using two different approaches. In the first case, presented in 5.2.1 the current steam system is analysed with the methodology shown in 4.2.2. The following assumptions are taken: Steam systems with current pressure levels and consumptions No changes of heat exchanger operation. Loads and utility levels are unchanged. Exchange of steam between plants is considered as a possible option.

In the second case, presented in 5.2.2, total site curves are developed using the grey box approach. The data for the heat exchangers with > 300 kW duty is represented with the gathered process data and additional steam consumption is represented in the TSA curves as the amount of utility consumed. The stream temperatures correspond to the utility temperature of the additional steam consumed (black box-approach). This analysis has the goal to identify: Modified utility system Changes of heat exchanger operation to recover maximum amount of heat. Introduction of new utilities and/or shifting of loads between utility levels.

In 5.3-5.5 the identified changes are discussed in terms of practical potential, cogeneration potential and finally a discussion about process streams below ambient temperature.

5.1 The current utility system

Heat to the processes is provided by steam, fuel gas, hot oil and hot water. Most of the heat is provided by steam and thus the analysis in 5.1.1 focuses on the steam system.

5.1.1

Steam production

An overview of steam production and use is given in Table 4. 73 % of the steam is generated from process heat and by incinerating by-products that must be incinerated on site. The rest has to be produced in boilers. All plants except AGA have boilers. Steam production is divided into three different groups: Steam from excess process heat Steam produced when cooling process streams with boiler feed water. The calculation is based on collected heat exchange data or available steam data. Steam from by-product incineration Some by-products have to be incinerated on site and are used for steam production. The fuels in this category cannot be sold or exported as fuel. Calculations are based on data from site visits. Steam from added fuel Added fuel is fuel produced on site Stenungsund (excluding the by-products above) or external fuel imported to Stenungsund. This steam production is calculated as the difference between the total steam production and the production from process heat + by-product incineration. Total steam production The sum of all steam production. 22

The steam production at different plants is presented in Table 4. Steam used in steam turbines has to be calculated separately to exclude the energy that is transformed into shaft power in the turbine. When steam turbines are used to drive process equipment, the available amount of steam is calculated as the steam exiting the turbine and being available for heating purposes. Thus the energy converted into mechanical energy is excluded. However, steam used in turbines for electricity cogeneration is not excluded and is considered available for other uses.

5.1.2

Steam use

The consumers of heat included in stream data (heat load >300 kW) uses 80 % of the steam produced. The exclusion of smaller steam demands has been necessary in order to handle the project within the given timeframe. Most of the heat demands are calculated from heat exchangers, but direct steam injection is also included. The fact that steam used for turbines which drive process equipment is excluded, limits the analysis since the current use of steam to turbines is accepted. For steam consumers <300 kW the steam demand is included in the analysis as described in 5.1.3. The data for the different plants at the site are listed in Table 4. Use in heat exchangers >300 kW - Taken from the collected data for users >300 kW. To estimate the steam use, the duty in the heat exchangers is used. Thus the exit pressure and temperature of the condensate is not known. Other use of steam Calculated as the difference of total steam production and steam consumed in heat exchangers collected in this study (>300 kW). In this case the steam condensate is assumed to be cooled to 120 oC for higher pressure steam (>2 bar(g) and 100 oC for low pressure steam (< 2 bar(g)).
Table 4 Total steam production, fuel demand and steam demand based on stream data and steam balances for the chemical cluster in Stenungsund
Steam production [MW] Excess process heat Steam from by-product incineration Steam from added fuel Total steam production Steam from added fuel Steam use [MW] Use in heat exchangers >300 kW Other use of steam

Site

AGA Akzo Nobel Borealis Cracker Borealis PE INEOS Perstorp Sum

26 141 35 14 25 241

6.4

0 25 32

0 57 173 41 39 79 389

42 % 18 % 4% 57 % 34 % 27 %

0 44 154 11 24 76 309

14 19 29 15 4 80

4.0 2.6 27 40

2 22 27 108

5.1.3

Handling of other steam users

Other users of steam steam not accounted for in the stream data gathered is calculated as the difference of total steam production and the steam use in heat exchangers >300 kW, as shown in Figure 10.

23

Excess process heat By-product incineration Steam from added fuel

241 MW 40 MW 108 MW Total steam production 389 MW

309 MW

Use in heat exchangers >300 kW Other use of steam

80 MW

Figure 10 Representation of the amount of steam from process and boilers, the total steam production and the amount of steam covered by stream data collected for this study.

It is important to include this steam consumption in the study. The temperatures of the process streams corresponding to the steam use are unknown. Hence, the process streams are represented at the temperature level of the steam used. This means that the utility level for these users is fixed and cannot be modified in the analysis. The use of steam for these small users is divided into use at different pressures according to steam balances of the different plants. The fraction of steam included in process stream data gathered in this study compared to the total steam production is 80 %. This is considered acceptable with the uncertainties involved in the data collection listed below. This indicates that a relevant heat load limit for including heat exchanger was chosen. Uncertainties in data collection: The exclusion of small heaters and other steam uses (tracing, heating, flares) Overall steam production and heat exchanger data are given at different times or for different operation circumstances Uncertainties in the heat exchanger stream data Uncertainties in steam measurement Work is extracted from the system as electricity and shaft power Steam excess that is not used for heating

5.2 Identifying improvements to the utility systems

5.2.1 Increased energy efficiency by integrating the existing steam systems

Since the steam system is a major part of the energy system the total sites steam systems are investigated more in detail. Steam considered in this section is only steam used to heat the processes and steam recovered from process cooling and by-product incineration. Steam that is produced in boilers by added fuel is not included. Figure 11 shows the particular excess or deficit of steam at each plant and pressure level. Steam consumed is subtracted from steam produced at each level. The numbers give the excess/deficit of steam at each pressure level.

24

49,8 -8,3 19,5 -12,0 -3,3 -3,0 -4,9 -1,9 -2,6 2,5 4,9 -4,4 -63,1 -15,3 -5,6 -8,6 23,9 8,3 -45,0 -31,4 -0,3 -7,5 -80,0

85bar(g) Borealis 40bar(g) Akzo 41bar(g) Perstorp 40bar(g) Borealis 28bar(g) Akzo 28bar(g) INEOS 20bar(g) Akzo 20bar(g) INEOS 20bar(g) Perstorp 14bar(g) Perstorp 11bar(g) Borealis 10bar(g) INEOS 8.8bar(g) Borealis 7bar(g) Perstorp 6bar(g) INEOS 6bar(g) Akzo 4bar(g) Borealis 2.7bar(g) Borealis 1.8bar(g) Borealis 2bar(g) Perstorp 1bar(g) Akzo 1bar(g) INEOS

Pressure level [bar(g)]

-60,0 -40,0 -20,0 0,0 20,0 40,0 60,0 Steam deficit [MW] Steam excess [MW]

Figure 11 Representation of the steam excess/deficit at the total sites pressure levels at the different plants (Steam produced in boilers by added fuel is not included)

The right side of the diagram shows excess of steam (in MW) at specific steam levels. The left side represents a deficit of steam at specific steam levels. This diagram enables to determine possible matches between steam levels with excess and deficit (excess of steam at a higher level can be transferred to a steam level with deficit). Therefore it gives guidance for determining exchange of steam across the total site. To find the ideal distribution of steam across the site, pinch analysis is used in form of a GCC. The GCC of a combined steam system (see Figure 12) shows a heating demand of 108.3 MW. The whole steam system is divided in temperature intervals. A surplus of steam at one can be transferred downwards the cascade to steam levels with heat deficit. Distribution of steam as suggested in the GCC gives a maximum utilisation of steam, a remaining steam deficit at the lowest possible steam levels and the highest cogeneration potential (shown in 5.3.3)

25

350 300 250 T (C) 200 150 100 50 0 0

Qadded=108.3 MW

Cogeneration with recovered steam

Cogeneration with steam from added fuel

50

100 Q (kW)

150

200

Figure 12 GCC of the steam system with minimum heating demand and pockets, showing potential for internal heat exchange

In order to realise the steam distribution as shown in the GCC eleven new steam headers have to be constructed. This leads to an ideal distribution of steam with deficits at the steam levels shown in Table 5.
Table 5 Steam levels and heat deficit at the total site with integrated steam systems

Steam level [bar(g)] 8.8 7 6 2 1 Sum

Deficit [MW] 26.8 15.3 14.2 44.2 7.8 108.3

The current steam systems are only partly connected, which means steam cannot be freely transferred between the companies as indicated in Figure 12. The current steam systems are described in chapter 7. The steam exchange between plants today: AGA receives 28 bar(g) steam from Akzo. Borealis PE receives 40 bar(g) steam from Borealis Cracker. Borealis Cracker will receive LP steam from Borealis PE.

Steam deficit at each plant can be determined with data from Figure 11. With the current steam system there is a deficit of steam at the steam levels listed in Table 6. This steam has to be produced by added fuel in the boilers.

26

Table 6 Steam levels with deficit if only current possibilities for steam exchange are considered

Remaining deficits Level bar(g) Load [MW] Akzo Akzo INEOS Akzo INEOS Borealis INEOS Borealis Akzo INEOS Perstorp Akzo INEOS Sum 40 28 28 20 20 8.8 10 1.8 6 6 2 1 1 8.3 3.3 3.0 4.9 1.9 25.3 4.4 7.9 8.6 5.6 27.3 0.3 7.5 108.3

The calculations for Table 6 are shown in Appendix 2. When comparing Table 5 and Table 6 it can be seen that the total steam demand doesnt change when combining the steam systems. This is due to the fact that the companies are in balance with their steam systems (none of the companies generates an excess of steam). The steam levels with a deficit in a combined system are lower, see Table 5. This gives a higher potential for cogeneration (29 MW instead of 10 MW) in a common steam system, as steam produced in the boilers at high pressure can be expanded to a lower steam level where it is needed. A common steam system also increase the potential and incentive for further energy recovery measures, e.g. steam recovered at one site which has no demand of steam at the recovered temperature can be transferred to another company if there is a demand. A common steam system for all steam levels would require an investment in 11 new steam headers between the plants. There are different specifications for boiler feed water at the plants, depending on the pressure in the boilers. Thus, an integrated steam system will also require investments in condensate treatment.

5.2.2

Using total site profiles to improve energy efficiency

5.2.2.1 Maximum theoretical potential for heat recovery In order to find the maximum theoretical potential for heat recovery from the processes the composite curves of the total site are plotted as shown in Figure 13. In this case it is assumed that heat exchange hypothetically occurs directly between the process streams. A temperature difference of 20 K is chosen. It can be seen that the ideal amount of heat recovery is 442 MW, which is all the heating demand of the processes. Additionally an excess of heat of 14 MW at high temperature levels can be generated. Direct heat exchange across the total site is very unrealistic. Therefore these values are only theoretical. Because of this other methods are developed, which

27

make it possible to represent the processes and their interaction with the utility systems. This enables to target for process integration using the utility system. 1600 1400 1200 Temperature [C] 1000 800 600 400 200 0 -200 0 200000 400000 600000 800000 1000000 1200000
T = 20 K

Qrec=442 MW

Qexcess=14 MW

Heat load [kW] Cold Composite Curve Hot Composite Curve

Figure 13 Composite curves of the total site in Stenungsund, representing the ideal potential for heat recovery by direct (process-process) heat integration

The method used is total site analysis (TSA) described in 4.2. First, curves are developed which represent the heat integration if no changes to the current heat exchangers are considered. This means the utility loads and levels are left untouched and only the utility systems of the different companies are assumed to be connected. The heating and cooling demand and the heat recovered are found in this way. The potential for cogeneration can be determined. In the next step the curves are used to identify ways to increase heat integration by changes to the heat exchangers. Changing the utility in certain heat exchangers in order to recover and use heat from the process in a better way. With help of the total site curves improvements to the utility system are identified which enable to approach the theoretical heat recovery shown in Figure 13. Site wide utility levels are set and new utilities are suggested. Several measures to recover energy are already implemented at the different plants. Heat is recovered directly (process-process heat exchange) and indirectly (heat recovery by the utility system). This study follows the grey box-approach (see 4.2.1) and therefore focuses on the process-utility interface for heat integration. Additional steam demand which is not covered by the streams gathered in this study is considered as black box. This is done as follows: The steam demand is included in the TSA curves with the amount of steam used for heating. It is included in the hot utility curve at the respective steam saturation temperature. 28

 In the sink profile with a temperature 10 K below the steam temperature. 5.2.2.2 Total site profiles (TSP) Figure 14 shows the TSP of the Stenungsund chemical cluster. The diagram represents cold and hot process streams (blue and red curve), and cold and hot utility curves (green and orange curve). Process cooling is shown at the left side of the graph. In order to withdraw heat from the hot process streams it is transferred to the cold utility curve. Process heating is shown at the right side. Hot utility is used to deliver heat to the cold process streams. Total cooling and heating demand, the amount of heat discharged to the atmosphere (to cooling water and air) and the heat recovered from the processes and supplied to the processes by different utilities can be determined. 1600 1400 1200 Temperature [C] 1000 800 600 Qatmos=560 MW Qrec=324 MW 400 200 0 -1200000 -800000 -400000 0 -200 Qcooling,total=953 MW Heat load [kW] Sink Profile Source Profile Hot Utility 400000
Qheating,total=442 MW

Cold Utility

Figure 14 TSP representing the process-utility interface of the Stenungsund chemical cluster

The total cooling demand of the processes is 953 MW. This cooling demand is covered by a number of utilities. Some of them are recovering heat from the processes as e.g. steam or hot water. Heat is also discharged to cooling water or to air. Processes that need to be cooled below ambient temperature utilise cooling media such as ethylene, propylene, cooled cooling water or ammonia. 324 MW heat is recovered of which 281 MW 4 is recovered as steam. The total amount of process heat discharged to the environment by CW and air is 560 MW. The rest of the cooling (77 MW) is done by cooling media.

Includes 40 MW of steam which is generated when by-products are incinerated. This incineration is necessary to reduce the environmental impact of the by-products or utilisation otherwise is not possible.

29

The total heating demand is 442 MW of which 389 MW is covered by steam at different pressure levels, while 55 MW is covered by hot water/oil, flue gases, refrigerant and steam condensate. Heating processes with refrigerant (29 MW) is done to recover cooling energy and thereby decrease the energy consumption of the refrigeration systems. 5.2.2.3 Total site composites (TSC) In order to determine the maximum heat integration at the total site with the current utility system the TSP are moved towards each other until the hot and cold utility curve intersect. This intersection point is the so called site pinch. It limits further heat integration by the utility system. This means that no further overlap of the curves is possible (Zhu & Vaideeswaran 2000). The overlap of the source and the sink profiles represents the amount of heat recovery by the utility system. The resulting graph can be seen in Figure 15. 1600 1400 1200 Temperature [C] 1000 800 600 400 200 0 -200 0 200000 400000 600000 800000 1000000 1200000 Site Pinch

Qheating=122 MW

Heat load [kW] Sink Profile Source Profile Hot Utility Cold Utility

Figure 15 TSC showing the maximum heat integration in the Stenungsund chemical cluster

The total site composites (TSC) show that with the utility system the heating demand of the total site is 122 MW. The theoretical savings potential if ideal process integration, as shown in Figure 13, is applied is: Excess of steam with ideal system + Current heating demand = Savings 14 MW (from Figure 13) + 122 MW (from Figure 15) = 136 MW Another finding from Figure 15 is that not all the heat recovered from the hot process streams is used to heat cold streams. This can be explained by: Some process streams deliver heat to the local district heating system, which is outside the system boundaries 30

Some circulating water systems dont utilise all the heat recovered and are therefore discharged to the cooling water system

The total amount of heat recovered in hot water systems at the different plants is 43 MW of which 9 MW are utilised. The difference of 34 MW can be explained by the factors mentioned above. The site pinch, which limits further integration within the utility system is at the 2 bar(g) steam level (132C). Elimination of this pinch point by changes in the utility system leads to additional energy recovery. The profiles shown in Figure 14 and Figure 15 are used to determine improvements to the utility systems. 5.2.2.4 Improvements to the utility system Introduction of a hot water circuit: One option to increase energy efficiency at the total site is to eliminate the site pinch by the introduction of a hot water circuit between 50 and 100C. Heat from hot process streams can be recovered in a circulating hot water system and delivered to cold process streams. This option is interesting because in the cluster a lot of excess heat is available in the temperature range chosen, which can be used to replace steam in several heat exchangers. By introducing such a hot water circuit steam is substituted with hot water. The demand of 2 bar(g) steam decreases which moves the pinch point so that the curves overlap can be increased. This can be seen in the enlarged section of Figure 16. The dotted line shows the current utility system. The green and the orange line represent the utility system after introduction of the hot water circuit. Introducing a hot water circuit results in: Recovery and utilisation of 60 MW of hot water between 50 and 100C Savings of 51 MW 5 steam at 2 bar(g) and below

Not all the 60 MW recovered in the hot water circuit are used to save low pressure steam. 9 MW of heat are already recovered and utilitsed in hot water systems. Therefore the hot water circuit saves 51 MW of steam at 2 bar(g) and below.

31

1600 1400 1200 Temperature [C] 1000 800 600 400 200 0 -200 0 200000

low pressure steam savings (51 MW) Hot water circuit (60 MW)

Qheating=71 MW

New Site Pinch

400000

600000

800000

1000000 1200000 Cold Utility

Sink Profile

Heat lead [kW] Source Profile Hot Utility

Figure 16 TSC after introduction of a hot water circuit between 50 and 100C

With integration of the hot water circuit the heating demand at the total site becomes 71 MW. The potential steam savings are 51 MW. An economic evaluation of the savings indicates annual steam savings of 122.4 MSEK/year 6. When introducing the hot water circuit a new site pinch is created at 151C. In order to further integrate the plants (up to 96 MW by transferring more heat in a hot water circuit) the site pinch has to be eliminated. This can be done by: Decreasing the heating steam levels to below 151C where it is possible or Increasing the level of recovered steam to above 151C if possible.

Utilise the potential for increased 2 bar(g) steam recovery: Another option to increase process integration with the utility system is to increase the recovery of 2 bar(g) steam. The total site profiles show that there is potential for recovery and use of 33 MW of 2 bar(g) steam. This can be seen in Figure 17. The enlarged section in the figure shows the potential for steam recovery. Comparing the current cold utility profile (dotted line) with the cold utility profile after increased 2 bar(g) steam recovery gives the amount of steam which can be additionally generated from process heat. Figure 13 also shows the effect of this measure on the total heating demand. Increased recovery of 2 bar(g) steam lowers the heating demand by added fuels to 89 MW, which corresponds to potential steam savings of 33 MW steam from added fuel.

Background to economic evaluation see Appendix 4.

32

An economic evaluation of this potential shows annual steam cost savings of 79.2 MSEK/year6. 1600 1400 1200 Temperature [C] 1000 800 600 400 200 0 -200 0 200000 600000 800000 1000000 1200000 Heat lead [kW] Source Profile Hot Utility Cold Utility 400000 Increased 2 bar(g) steam generation (33 MW)
Qheating=89 MW

Sink Profile

Figure 17 TSC with increased 2 bar(g) steam recovery

It can be seen in Table 6 that almost all the current deficit of steam at 2 bar(g) and below (35.1 MW) can be covered by this measure. Most of the 2 bar(g) steam that could be generated from process heat comes from Borealis. At the Borealis site there is currently no deficit of steam at this pressure level. The steam generated at Borealis could be used to satisfy the demand at Perstorp, INEOS and Akzo Nobel (see Table 6). 51 MW of 2 bar(g) steam can be saved before an excess of 2 bar(g) is generated. Further utility savings can be reached by shifting loads in heat exchangers to lower levels or by recovering steam from the processes at higher levels. Further process integration with the utility system In order be able to utilise all heat recovered that can be recovered in a hot water system and the increased amount of 2 bar(g) steam it is necessary to do further adjustments to the utility system. When moving the total site profiles towards each other, new pinch points are created. They can be eliminated by adjusting utility levels (shifting loads between utilities by e.g. increasing temperature of steam recovery and/or lower temperature of steam used for heating). Figure 18 shows the TSP after several suggested changes to the utility system.

33

400 4 Temperature [C] 3 2 1 200 100 0 -800000 -400000 -100 -200 Heat load [kW] Sink Profile Cold Utility Source Profile Current cold utility Hot Utility Current hot utility 0 400000 300

6 5

-1200000

Figure 18 TSP showing measures to reach maximised energy recovery by the utility system

Changes in both the process heating and cooling have to be applied to further improve the process integration at the total site. E.g. if the process is cooled by steam generation there also has to be a demand for the steam in the process. Otherwise the steam cannot be utilised and has to be discharged. Adjustments to the process heat exchangers have to be done to make the steam generated in the process available to them. Table 7 summarises the measures shown in Figure 18 and their effects on the heating/cooling demand.
Table 7 Measures and effects of changes to the utility system shown in Figure 18. NB Measures 3-6 is not described in text above.

No. 1 2 3 4 5 6

Measure Introducing a circulating hot water system between 50 and 100C Increase LP steam recovery Increased 8.8 bar(g) steam recovery by 50 MW Increased steam recovery at 85 bar(g) by 44 MW Lower steam level for process heating by 129 MW Replace flue gas by steam by 10 MW

Advantages Saves 96 MW 7 of LP steam and cooling demand Increase in LP steam by 33 MW covers LP steam demand Necessary for further integration (to cover 8.8 bar(g) demand) Necessary for further integration Necessary for further integration (utilisation of recovered steam) Necessary for further integration

96 MW is the potential for heating process streams with hot water in a range between 50 and 100 C

34

Assuming that all the suggested changes are applied the potential for increased heat recovery with the suggested measures is 129 MW. This includes steam savings of 122 MW of the plants added fuel demand and additional surplus of 7 MW of steam at a pressure of 85 bar(g). With the adjustments described in Table 7 the utility curves overlap as shown in Figure 19. 7 MW excess steam realised by the changes suggested should be compared to the theoretical steam excess of 14 MW shown in Figure 13. It can be seen that with the suggested adjustments to the utility system it is possible to almost reach the efficiency of the ideal system. The surplus of steams can be used for e.g. electricity production or to supply new processes with heat. The TSC representing this utility system to maximise energy recovery is presented in Figure 19. The utility levels for an optimised utility system can be found in the TSC. Here it is suggested: 1600 1400 1200 Temperature [C] 1000 800 600 400 200 0 -200 0 200000 Sink Profile 400000 600000 800000 1000000 1200000 Heat lead [kW] Source Profile Hot Utility Cold Utility Circulating hot water between 50 and 100C 2 bar(g) steam 8.8 bar(g) steam 40 bar(g) steam 85 bar(g) steam Qsurplus=7 MW of 85 bar(g) steam

Figure 19 TSC for utility system modifications that maximise heat recovery

A table with loads and pressure levels of the utility system shown in Figure 19 is given in Appendix 3. Effects on cooling water demand If heat from the process is recovered and used instead of being discharged the cooling demand is decreased. This is especially important for plants using fresh water, as the 35

cooling water use is restricted for them. But it also lowers the pump work in the cooling water pumps and therefore saves electricity.

5.3 Qualitative evaluation of potential solutions

The practical potential to implement the modifications suggested in 5.2 is evaluated qualitatively. This is done by a qualitative assessment of utility changes in each heat exchanger that are included in the suggested changes. The qualitative feasibility analysis is performed according to 4.2.7. In discussions with process engineers at the site complexity of utility changes is assessed and heat exchangers where it is impossible to use another utility are identified. The following grading system is applied: A. possible: no process constraints, moderate investment costs are expected B. possible, with constraints: several or more complex changes are necessary, higher investment costs are expected C. impossible: process constraints make changes impossible The results are shown in: Figure 20, for the suggested hot water circuit classified in category A Figure 21, for the suggested hot water circuit classified in category A and B Figure 22, for increase 2 bar(g) steam generation from process heat classified in category A and B

This estimation was done in a screening session and does not involve any technical or economical calculations. But the results give an indication of how realistic the theoretical potential is. The following figures visualise the practical potential for implementing utility systems as suggested in 5.2.2.

5.3.1

Hot water circuit

45833

50000 45000 40000 35000 30000 25000 20000 15000 10000 5000 0

Heat load [kW]

30038 25883

Heating with circulating hot water

11336 12849 6157

Cooling with circulating hot water

4755 2000

260

992

Akzo Borealis Borealis Ineos Perstorp PE Cr

Figure 20 Practical potential for implementing a hot water circuit considering changes classified as A

36

Figure 20 shows the practical potential for implementing a hot water circuit. Only heat sources and sinks in category A are included. It can be seen that all the plants have a high potential to deliver heat to a common hot water circuit. The highest demand for heat from such a circuit is found at Borealis cracker plant, Perstorp and Akzo. Total heat delivered to hot water circuit in category A: 85 MW Total heat withdrawn from hot water circuit in category A: 55.2 MW

Utilising the potential of a hot water circuit in category A, steam savings of 132.5 MSEK/year6 can be realised. Figure 21 shows the practical potential for implementing a hot water circuit for category A and B. All plants could deliver a large amount of heat to the hot water circuit. By far the highest heat sink for a hot water circuit is the cracker plant but also the other plants have considerable heating demand between 50 and 100C. Total heat delivered to hot water circuit in category A + B: 255 MW Total heat withdrawn from hot water circuit in category A + B: 83.5 MW

Utilising the potential of a hot water circuit in category A and B, steam savings of 200.4 MSEK/year6 can be realised. 100000 90000 80000 70000 60000 50000 40000 30000 20000 10000 0
95845 76222 68072 64303

Heat load [kW]

Heating with circulating hot water Cooling with circulating hot water
6157

13501 4747 1793 2778

Akzo Borealis Borealis Ineos Perstorp PE Cr

Figure 21 Practical potential for implementing a hot water circuit considering changes classified as A and B

Practical consequences of the introduction of a hot water circuit: The introduction of a hot water circuit implies that process coolers have to be redesigned to use hot water instead of cooling water or air. Process heaters have to be redesigned for hot water instead of steam heating. Hot water pipes between several plants have to be constructed, as most of the consumers of heat are situated at the cracker site and at Perstorp but the sources are spread out across the cluster. Difficulties when integrating several plants with a hot water circuit: Different operating times Long distance between the plants Larger area in the new heat exchangers expensive in congested areas 37

Many heat sources and sinks temperature regulation difficult Utilisation of saved steam can involve further investments 8

Potential for supplying hot water to the district heating system: Borealis and Perstorp deliver heat to the district heating system of Stenungsund but the demand for district heat in Stenungsund is limited. It has been discussed to extend the Gothenburg district heating system to Stenungsund. This gives the opportunity for the cluster to deliver more heat to the system. Not all the potential for heat recovered can be utilised in the cluster and also in some cases it might be more profitable to deliver waste heat to the district heating system instead of using it to replace steam. This is especially the case when more complex retrofits are necessary to replace steam with hot water in the processes.

5.3.2

Increased steam recovery

Figure 22 shows the potential for increased 2 bar(g) steam generation. A large amount of steam can be produced at the cracker plant. This steam could be generated by further cooling the flue gases of the older cracking furnaces. Other potential sources of low pressure steam are available at Borealis PE plant and at INEOS. 25000
22410

20000 Heat load [kW] 15000 10000 5000 0 Akzo Borealis PE Borealis Cr Ineos Perstorp
3550 690

Category A Category B

Figure 22 Practical potential for increased generation of 2 bar(g) steam generation

The total practical potential for 2 bar(g) steam generation in category A and B is 26.6 MW, compared to the theoretical value of 33 MW. An economic evaluation shows potential savings of 10 MSEK/year for category A measures and 64 MSEK/year6 if both A and B measures are implemented. Increased 2 bar(g) recovery demands the installation of steam generators and the construction of steam pipes from Borealis to Perstorp and INEOS, as most of the

If steam savings apply at a company that currently has no demand at the this level, further investments to utilise the steam are necessary. E.g. heat exchangers that work on the sam steam level or piping to transfer the steam to another plant.

38

potential steam sources are located at Borealis but Perstorp and INEOS have a demand for 2 bar(g) steam

5.3.3

Further changes to the utility system to increase savings

To avoid an excess of low pressure steam, the low pressure steam demand must be increased. According to the total site profiles (see Figure 14) and the stream data gathered it is possible to decrease the steam pressure in several heat exchangers. Besides increased heat integration, using steam at lower pressure increases the potential for cogeneration. The potential to use steam at lower pressure to heat processes is 26.2 MW with moderate 34.8 MW including more complex changes.

5.4 Cogeneration potential

Four cases of electricity generation are estimated. Theoretical cogeneration potential with the current utility levels and current steam demand and excess in an integrated steam system All steam is produced at Borealis and distributed to plants with a deficit of steam below 8.8 bar(g). Cogeneration potential with a totally integrated steam system and a new, common boiler Cogeneration to produce steam at 14 bar(g) and 2 bar(g) at Perstorp

5.4.1

Theoretical cogeneration potential with the current utility levels and current steam demand and excess in an integrated steam system

The cogeneration potential for an entirely connected steam system is shown in Figure 12. The red arrows represent the cogeneration potential with steam recovered from the process. The blue arrows show steam produced from added fuel in the boilers. For the theoretical case it is assumed that steam from the process can be expanded in turbine from the pressure at which it is produced down to the pressure level at which it is used. Steam from the boilers can be expanded from generation pressure (here 85 bar(g) is assumed) to its utilisation pressure to produce electricity. The theoretical cogeneration potential is 29 MW if steam is expanded as shown in Figure 12. Today 10 MW electricity is produced. The additional 19 MW corresponds to 40 MSEK/year6.

5.4.2

All steam is produced at Borealis and distributed to plants with a deficit of steam below 8.8 bar(g).

The data for the following analysis is extracted from Table 6. An option for increased cogeneration with the existing cogeneration equipment and utility system is described more in detail. Borealis has the possibility for cogeneration when producing 85 bar(g) steam and expanding it to 8.8 bar(g). Therefore it can be advantageous to cover the deficits of steam at levels below 8.8 bar (g) at other sites by steam from Borealis. The deficits of steam at other sites are shown in Table 8. These deficits could be covered by steam from Borealis and thereby electricity is produced.

39

Table 8 Companies with deficit at steam levels below 8.8 bar(g) which could be additionally supplied with cogeneration steam from the cracker plant

Plant Akzo Akzo Perstorp INEOS INEOS Sum

Level [bar(g)] Load [MW] 6 1 2 6 1 8.6 0.3 27.3 5.6 7.5 49.3

Necessary measures: Borealis to INEOS: 13.1 MW to cover deficit of 6 and 1 bar(g) steam Borealis to Akzo: 8.9 MW to cover deficit of 6 and 1 bar(g) steam Borealis to Perstorp: 27.3 MW to cover deficit of 2 bar(g) steam Improved condensate treatment Covers INEOS, Akzos and Perstorps demand of steam below 8.8 bar(g). Additional 49.3 MW of 8.8 bar(g) steam must be produced by cogeneration at Borealis. Production of additional 8.6 MW electricity when expanding steam from 85 to 8.8 bar(g). Additional 59 MW 85 bar(g) steam must be produced at Borealis.

Effects:

An economic evaluation of the suggested measure indicates a revenue increase of 18 MSEK/year.

5.4.3

Cogeneration potential with a totally integrated steam system and a new, common boiler

Integration of the total site utility system aims at reducing energy use at the total site. The results of the TSA analysis show that no steam will have to be produced with added (purchased) fuel if all suggested measures are implemented. However, a more realistic result of a modification is that there will still be a demand for steam production. Here, the cogeneration potential in a new separate, common boiler with a backpressure steam turbine is assessed. The boiler is assumed to produce steam at 140 bar and 540 oC. Depending on what measures are carried out in the processes the remaining steam demand will be at different steam pressures, thus evaluation is carried out for two different back-pressure levels.

40

70 60

30

MW or MSEK

50 40 30 20 10 0 0

MW or MSEK

Elctricity production [MW] Revenue [MSEK/year]

25 20 15 10 5 0

Elctricity production [MW] Revenue [MSEK/year]

20

40

60

80

100

120

20

40

60

80

100

120

Steam from added fuel [MW] Back pressure 2 bar(g)

Steam from added fuel [MW] Back pressure 40 bar(g)

Figure 23 Electricity production and revenue (electricity sold - extra fuel) for backpressure 2 bar(g) and 40 bar(g).

It can be seen in Figure 23 that with increased energy efficiency (= lower steam demand), the cogeneration potential becomes less. This is due to the fact that less steam has to be produced in the boilers at high pressure which takes away the opportunity to expand this steam in a turbine to lower levels where it is applied in the process. Figure 23 also show that the cogeneration potential increases if the remaining steam demand is at a low pressure.

5.4.4

Cogeneration at Perstorp

Perstorp has a high steam demand at low pressure and generates steam at 40 bar(g). A steam turbine with backpressures at 14 bar(g) and 2 bar(g) could generate 8.5 MW electricity (corresponding to a revenue of 16 MSEK/year6).

5.5 Heat exchangers operating below ambient temperature

All plants except Perstorp are operating cooling systems to provide cooling below the temperature that the cooling water system can provide. At the AGA air separation process the process streams are to a great extent integrated and utility is not used with a few exceptions. Most cooling utility is used at the Borealis Cracker plant, and an analysis of the system is included in a Master Thesis (Hedstrm and Johansson, 2008). Cooling below ambient is expensive and the cost increases sharply with lower temperature (Kemp 2007). Reduction in cooling demand by increased process integration will save energy, electricity or steam depending on how the compressors are driven. Since cooling is expensive, the cooling utility levels are carefully chosen to fit the temperature level. However, heating of cold streams below ambient temperature is in some cases applied with utility far above the necessary temperature level. Except for saving high temperature utility, such as steam, a more careful design could use the cooling capacity of the sub-ambient stream that has to be heated to reduce the energy use in cooling systems. To transfer cooling energy from streams below ambient temperature to streams that have to be cooled to low temperatures, so called low temperature heat transfer media can be used. Depending on the temperature this can be e.g.: 41

Sodium chloride brine: down to -21 oC Calcium chloride brine: down to 40 oC Ammonia Liquid ethylene or propylene: even lower temperatures (Heaton 1996, p.244)
Load [MW] 2 1 3.5 Tstart [C] -103 -135 Ttarget [C] 40 20

Table 9 Streams heated by steam at low temperatures Steam description Ethylene vaporizers at the EFAB Oxygen and nitrogen vaporizers at AGA (intermittent) Vaporizers at Borealis operating below 0 C tank9:

Potential low temperature sources are shown in Table 9. It shows streams below ambient temperature that are currently heated with steam. Utilising them to recover cooling energy can save both steam and power used to drive the cooling systems. The gathered stream data indicates that 29 MW of heat is transferred from the refrigerant to cold streams in order to recover some of the energy invested in cooling. Some of that energy is recovered at higher temperatures and not included in the figure above. If the cold streams currently heated by steam instead are heated with process streams that have to be cooled, cooling demand at low temperatures can be reduced which will result in substantial energy savings. The energy requirement in a refrigeration system depends on the temperature:
Table 10 Power demand of cooling energy depending on the refrigeration temperature (Kemp 2007)

Refrigeration temperature [C] Power use [MWel/MWcooling] -40 -100 0.6 1.6

Table 10 shows the power demand of cooling energy depending on the refrigeration temperature. The shaft work for the compressor is either generated with electricity or steam. By utilising the cooling energy at EFAB and AGA (3 MW, see Table 9) refrigerant at -100 oC can be replaced. This can save 4.8 MWel, and 3 MW steam, corresponding to 30 MSEK/year. If the vaporizers at Borealis can replace refrigerant at -40 oC (3.5 MW, see Table 9), 2.1 MWel and 3.5 MW steam corresponding to 18 MSEK/year can be saved. The steam savings above are included in the savings discussed in 5.3. Further integration of the cooling systems of the different plants might have benefits other than energy savings, such as cost reduction for operation and maintenance of cooling system equipment.

Evaporation of ethylene from the EFAB tank is expected to increase.

42

6 Conclusions
Total site analysis (TSA) is based on analysis of current use of utilities for heating and cooling and the focus is on finding energy saving potential by integration of the utility systems for the included plants. First stream data was collected for the most significant process streams and completed by data on utility consumption for other uses. Thereby all the utility demand was accounted for in the study. Starting with the analysis of the current steam system, potential improvements were identified. After that additional options for increased energy savings were developed by also allowing changes in heat exchanger operation. The results of the study reflect the current heat energy situation at the total site and give suggestions on how fuel and cooling capacity savings as well as the potential for cogeneration can be utilised. In order to assess the feasibility of the heat integration measures their practical potential was estimated. Present energy system Total heating demand for the total site today is 442 MW. Most of the heating demand is covered with heat recovered from the processes. 122 MW heat has to be supplied as added fuel, directly or as steam/hot oil. Assuming a boiler efficiency of 80 %, 150 MW fuel is required. This corresponds to operating costs of approximately 288 MSEK/year. Most of this fuel is produced as a by-product from the cracker plant, but approximately 25 MW fuel has to be imported to Stenungsund. Theoretical potential for reduced fuel usage The TSA analysis shows that with an optimal utility system all heating demand supplied with added fuel can be saved and there will be a heat surplus of 14 MW. The total savings are then 122 MW + 14 MW = 136 MW. Measures required to achieve theoretical potential Introduction of a hot water belt that recovers heat at 50-100 C and replaces steam in heat exchangers that operate below 90 C. This can save 51 MW of steam. More steam savings can be achieved when changing heat exchanger operations so that steam replaced by hot water can be utilised to heat other processes. In this case the theoretical potential for steam utility savings is 96 MW. Increased low pressure steam generation from process streams that are today cooled by cooling water or air can increase LP steam generation by 33 MW. If all these measures could be implemented the energy savings would be 129 MW. Practical potential of increased steam recovery and introduction of a hot water circuit Implementation of a hot water circuit shows practical potential steam savings of 55.2 MW (132 MSEK/year) with moderate changes (83.5 MW including more complex changes, 200 MSEK/year). Technically the introduction of a hot water circuit includes hot water pipes between several plants, as most of the consumers of heat are situated at the cracker site and at Perstorp but the sources are spread out across the cluster. 43

The practical potential for increased 2 bar(g) steam recovery is estimated to 4.2 MW (10 MSEK/year) with moderate changes and 26.6 MW including more complex changes (64 MSEK/year). Increased 2 bar(g) recovery implies the construction of steam pipes from Borealis to Perstorp and INEOS, as most of the potential steam sources are located at Borealis but Perstorp and INEOS have a demand for 2 bar(g) steam. Cogeneration The potential for cogeneration will decrease when the steam demand is reduced. If all steam demand below 8.8 bar(g) is produced at Borealis at 85 bar(g)it is possible to cogenerate 8.6 MW electricity which corresponds to a revenue increase of 18 MSEK/year. This measure requires new steam headers from Borealis cracker to Akzo, INEOS and Perstorp plus additional condensate treatment. Complete integration of all the steam systems of the 5 companies and no changes to the heat exchangers utility supply, results in an increased cogeneration potential of 19 MW (app. revenue is 40 MSEK/year). An option with a new, common boiler connected to a back-pressure steam turbine was evaluated. The cogeneration potential depends on the pressure level on the exiting steam and on the steam demand and is estimated to between 0-40 MW (0-82 MSEK/year). Cooling below ambient Some process flows below ambient are heated with steam. If they instead were heated with process flows that require cooling media, both electricity and steam would be saved.

44

7 Future Work
The work presented gives a base for future energy efficiency projects in the chemical cluster in Stenungsund. Further technical and economical analysis on the integration possibilities by a circulating hot water circuit should be investigated. Therefore the qualitative analysis performed in 5.3 can be used to identify the most interesting options and evaluate them according their technical and economical potential. Process integration of streams below ambient temperature should be analysed more in detail. Savings in cooling loads can be of high economical interest as cooling below ambient temperature is very energy demanding. The work accomplished in this study has also underlined a need for further development of the TSA methodology so as to enable better analysis of complex sites where steam is exchanged between plants and different pressure levels and used as reactant, utility, mechanical shaft work driver, and cogeneration. Further complications arise from the fact that steam drivers are used to drive high pressure compressors for sub-ambient refrigeration systems, thus creating a physical link between the heating and cooling utility systems. Further development of the methodology is one of the major components of suggested future work. Another suggestion for future work is to use the results from this study to investigate opportunities to integrate energy-intensive climate-friendly processes such as advanced biorefinery concepts or carbon capture and storage in the chemical cluster in Stenungsund.

45

8 References
Akzo Nobel, 2004. A site with a view - Site Stenungsund. Available at: www.ethyleneamines.com/NR/rdonlyres/A7FAF030-9E84-417A-AA10CBE70AC3444B/0/Akzo_eng_LR.pdf [Accessed February 18, 2010]. Allsopp W., M. & Vianello, G., 2009. Poly(Vinyl Chloride) : Ullmann's Encyclopedia of Industrial Chemistry. In Ullmann's Encyclopedia of Industrial Chemistry. Weinheim: Wiley & Sons. Available at: mrw.interscience.wiley.com/emrw/9783527306732/ueic/article/a21_717/current/ht ml?hd=All,pvc&hd=All,process [Accessed September 30, 2009]. Bagajewicz, Miguel, and Hernn Rodera. 2001. On the use of heat belts for energy integration across many plants in the total site. The Canadian Journal of Chemical Engineering 79, no. 4: 633-642. doi:10.1002/cjce.5450790424. Borealis AB, 2009. Borealis Miljrapport 2008, Borealis AB. Available at: www.borealisgroup.com/pdf/stenungsund/Borealis_Miljorapport_2008_1.pdf [Accessed September 22, 2009]. Brown, S.M. 1999. The drive for refinery energy efficiency. Petroleum Technology Quaterly (Autumn) Franck, Per-ke. 1997. Lecutre notes: Pinch technology in practice. Chalmers Industriteknik Energiteknisk Analys. Hackl, Roman, and Stephanie Perret. 2009. Power production with Organic Rankine Cycle technology utilizing waste heat from a cracker and three polyethylene units. Master Thesis. Gteborg: Chalmers University of Technology. http://publications.lib.chalmers.se/cpl/record/index.xsql?pubid=94982. Heaton, C.A., 1996. An Introduction to industrial chemistry, Springer. Hedstrm, Hanna, and Daniella Johansson. 2008. A Pinch targeting study of the Borealis Cracker plant in Stenungsund. Chalmers university of technology. Master Thesis. Gteborg: Chalmers University of Technology. http://publications.lib.chalmers.se/cpl/record/index.xsql?pubid=68102 Josefsson, Lars. 2009. INEOS Extern miljredovisning EMAS 2008. INEOS ChlorVinyls, May 12. http://www.emas.se/redovisningar/emas146-2008se.pdf. Kemp, Ian C. 2007. Pinch analysis and process integration. Butterworth-Heinemann. Klemes, J., V. R. Dhole, K. Raissi, S. J. Perry, and L. Puigjaner. 1997. Targeting and design methodology for reduction of fuel, power and CO2 on total sites. Applied Thermal Engineering 17, no. 8-10 (August): 993-1003. doi:10.1016/S13594311(96)00087-7. Kreith, Frank, and D. Yogi Goswami. 2007. Handbook of energy efficiency and renewable energy. CRC Press. Linhoff, Bodo, and Vimal Sahdev. 2002. Ullmanns Encyclopedia of Industrial Chemistry: Pinch Technology. In Ullmanns Encyclopedia of Industrial Chemistry. Weinheim: Wiley VHC. Linnhoff March. 2000. The Methodology and Benefits of Total Site Pinch Analysis. TotalSiteMethodology. www.linnhoffmarch.com/pdfs/TotalSiteMethodology.pdf. Meyers, Robert. 2004. Handbook of Petrochemicals Production Processes. 1st ed. McGraw-Hill Professional, April 15. 46

Perry, S., Klemes, J. & Bulatov, I., 2008. Integrating waste and renewable energy to reduce the carbon footprint of locally integrated energy sectors. Energy, 33(10), 1489-1497. Pollard, MilosLetter. 2010. Allmnt om Site Stenungsund. February 26. Raissi, K. 1994. Total site integration. PhD Thesis, University of Manchester Institute of Science and Technology. Smith, R., 2005. Chemical process design and integration, Chichester West Sussex England ;Hoboken NJ: Wiley. Spetz, R. et al., 2006. Energy survey: Borealis PE-site 2004. UIGI. 2009. Universal Industrial Gases, Inc.Cryogenic Air Separation and Liquefier Systems. www.uigi.com/cryodist.html. Whiteley, S. 2000. Polyolefins : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience. In Ullmann's Encyclopedia of Industrial Chemistry. 7th ed. Weinheim: Wiley VHC, June 15. mrw.interscience.wiley.com/emrw/9783527306732/ueic/article/a21_487/current/ab stract. Zhu, Frank X. X., and Lakshmi Vaideeswaran. 2000. Recent research development of process integration in analysis and optimisation of energy systems. Applied Thermal Engineering 20, no. 15-16 (October 1): 1381-1392. doi:10.1016/S13594311(00)00013-2.

47

9 Appendix

9.1 Appendix 1
Process flowsheets

Figure 24 Process flowsheet of an air separation unit (UIGI 2009)

Figure 25 The Ethylene oxide process at Akzo Nobel (Akzo Nobel 2004)

Figure 26 The Alkoxylation process at Akzo Nobel (Akzo Nobel 2004)

Figure 27 The amines process at Akzo Nobel (Akzo Nobel 2004)

HXTR-Train E1651 A-H/ E1656 A-H

3-stage compressor C-1701/02 Distillate drum D-1681 Water to sour water stripper Caustic wash tower T-1702 Demethanizer T-1801X Dryers with molecular sieves

Cracking furnaces T=850C

Ethane Propane Butane Naphta

Primary fractionator

E-1654 A/B Fuel oil to storage

Vapor Ethylene product

C2 splitter Ethylene Product Ethane recycling Depropanizer T-1703 T-1806/ 1807 Acetylene reactors R-1801 A-B C-4, to SHP/ETBE or storage

Ethane recycled back to cracking furnaces

Ethylene/Ethane Splitter T-1803 Deethanizer T-1802

Debutanizer T-1704

C-5 and heavier, to SCN storage

Propylene Product

Ethane/ light HC

Propylene Reactors R-1802 A-C C-3 back to Cracking Propylene/ Furnaces Propane Splitter T-1805 Secondary deethanizer T-1809 Propylene rerun Tower T-1804

Figure 28 The cracker process at Borealis (Hackl & Perret 2009)

2 3 4 6 1
Catalyst

7 5

1. 2. 3. 4.

Gas purification Polymerisation Separation Intermediate storage

5. Hot compounding 6. Blending 7. Packaging Storage, Distribution

Figure 29 The low pressure PE process at Borealis (Borealis AB 2009)

Gas phase reactor Loop reactor Catalyst Prepolymerisation reactor Gas

Flash

Diluent recycle

Polymer Degassing Pelletising Polymer

Ethylene Comonomer Hydrogen

PE

Figure 30 The PE3 Borstar process at Borealis (Borealis AB 2009)

Figure 31 The new high pressure PE process at Borealis (Meyers 2004)

Figure 32 The chlorine process at INEOS (Josefsson 2009)

Figure 33The VCM process at INEOS (Josefsson 2009)

Figure 34 The PVC process at INEOS (Josefsson 2009)

9.2 Appendix 2
Table 11 shows how the excess or deficit of steam at certain steam levels in the current utility systems was calculated. Excess of steam at a higher steam level can be used to cover a deficit at a lower level.
Table 11 Calculation of steam levels with deficit

Steam level 85bar(g) Borealis 40bar(g) Borealis 8.8bar(g) Borealis 4bar(g) Borealis 2.7bar(g) Borealis 1.8bar(g) Borealis Sum

Excess/Deficit [MW] 49,8 -12,0 -63,1 28,8 8,3 -45,0

Deficit after let down [MW]

-25.3

-7.9 -33.2

40bar(g) Akzo 28bar(g) Akzo 20bar(g) Akzo 6bar(g) Akzo 1bar(g) Akzo Sum

-8,3 -3,3 -4,9 -8,6 -0,3

-8.3 -3.3 -4.9 -8.6 -0.3 -22.5

41bar(g) Perstorp 20bar(g) Perstorp 14bar(g) Perstorp 7bar(g) Perstorp 2bar(g) Perstorp Sum

19,5 -2,6 2,5 -15,3 -31,4 -27.3 -27.3

28bar(g) INEOS 20bar(g) INEOS 10bar(g) INEOS 6bar(g) INEOS 1bar(g) INEOS Sum

-3,0 -1,9 -4,4 -5,6 -7,5

-3.0 -1.9 -4.4 -5.6 -7.5 -22.3

9.3 Appendix 3
Table 12 Loads and pressure levels of the utility system for maximised total site integration; including generation and consumption of steam in the improved system

Steam level [bar(g)] 85 bar(g) 40 bar(g) 20 bar(g) 14 bar(g) 11 bar(g) 10 bar(g) 8.8 bar(g) 2.7 bar(g) 2 bar(g) 1 bar(g) HotW Sum

Generation [MW] 51 42 29 15 9 13 53 21 80 1 107 421

Consumption [MW] 1 69

132 105 107 414

Difference [MW] 50 -27 29 15 9 13 -79 21 -25 1 0 7

9.4 Appendix 4
Table 13 Prices for steam, electricity and fuel assumed for economical evaluation of energy efficiency measures SEK/MWh Assumed prices: Saved steam Electricity Additional Fuel for steam production 300 600 240

Table 14 Economical evaluation ofenergy efficiency measures suggested in the report with an assumed running time of 8000 h/year Electricity Steam Addition produced/ Value Incomes Cases savings al fuel saved [SEK/h] [MSEK/y] [MW] [MW] [MW] Cogeneration by sending steam from Borealis to 8,6 13,6 1890,3 15,1 INEOS,Akzo,Pertorp, see 5.3.3 Introducing only hotW, see 5.2.2 Increased steam recovery, see 5.2.2 Maximised heat recovery (hotW, LP-steam and steam level adjustments), see 5.2.2 hotW category A, see 5.3 hotW category A + B, see 5.3 Steam category A, see 5.3 Steam category A + B, see 5.3 Common steam system, see 5.3.3 Electricity from 7 MW excess steam Cogeneration at Perstorp (see 5.4.4) Utilising cooling energy at EFAB and AGA (see 5.5) Utilising low-T Borealis (see 5.5) evaporators at 2,1 1260,0 10,1 51 33 15300,0 9900,0 122,4 79,2

122 55,2 83,5 4,24 26,6

2,3

37980,0 16560,0 25050,0 1272,0 7980,0

303,8 132,5 200,4 10,2 63,8 39,7 11,0 16,3

19,0 2,3 8,5

26,8

4959,0 1380,0

11,9

2040,0

4,8

2880,0

23,0