• No results found

Flue gas heat recovery for district heating : Analysis of flue gas condensation on a crematorium facility at Hovdestalund, Västerås.

N/A
N/A
Protected

Academic year: 2021

Share "Flue gas heat recovery for district heating : Analysis of flue gas condensation on a crematorium facility at Hovdestalund, Västerås."

Copied!
47
0
0

Loading.... (view fulltext now)

Full text

(1)

FLUE GAS HEAT RECOVERY FOR

DISTRICT HEATING

Analysis of flue gas condensation on a crematorium facility at Hovdestalund,

Västerås.

CHARLIE LINDQUIST

School of Business, Society and Engineering Course: Degree project work energy engineering Course code: ERA401

Credits: 30 hp

Program: Master’s program in sustainable energy systems

Supervisor: Anders Avelin Examiner: Monica Odlare Costumer: Mälarenergi Date: 2021-06-08 Email:

(2)

ABSTRACT

This degree project investigates the technical and economic possibility of a heat recovery system on Hovdestalund crematorium. The crematorium will have a solution that allows the facility to become both a producer and consumer to the district heating network. This work became relevant because the Swedish church is going to expand Hovdestalund with

additional ovens and therefore energy efficiency is in their sight. With the help of

Mälarenergi, a business model to be connected to the district heating network is created for Hovdestalund. The project will go through how to handle the flue gases, either by connecting the crematorium to the district heating network, or to cool down the gases with a cooling tower. When using the flue gases for district heating, two cases of different heat recovery potential are investigated. The first case is taken from an already made investigation by Kadesjös on Hovdestalund where they assume a potential of 400kWh/cremation. The second case is taken from literature study where the potential is 242,5kWh/cremation.

A technical solution that can separate when heat is supplied to the network versus withdrawn is chosen to help with billing and general surveillance across the system. Connecting

Hovdestalund to the network will not affect Mälarenergi’s operation negatively, the temperature can reach around 95 degrees Celsius and thus considered as high-grade heat. Having a lower temperature supplied to the network at a higher amount could lead to some problems in lowering the networks overall temperature. Using a cooling tower would not be as optimal as it would only cool down the flue gases, there would be no self-consumption nor heat sold to the network. Most of the investment cost will come from laying down a new pipeline for the district heating. The pipeline will have a length of roughly 140m which results in a cost of ~1MSEK. Along with the district heating unit, the total investment cost becomes 1 138 000 SEK. To investigate the economic possibility, the net present value is used. The systems profit comes from selling excess heat or the value of self-consuming heat rather than buying it from Mälarenergi. Between the two recovery potential cases, bought heat, self-consumed heat, and heat demand will be the same as the only difference will be the amount of sold heat. These will be the same because during operation time of the ovens, the

recovered heat will cover the hourly demand entirely. In case 1, the amount of sold heat will result in ~90tSEK annually which will lead to a payback time of 6-7 years. Case 2 will sell heat for ~45tSEK annually, this will lead to a payback time of 8-9 years. Compared to

Kadesjös result which received a final cost of 10MSEK after 20 years, this report shows a final cost of 7,4MSEK and 6,49MSEK based on recovery potential.

Keywords: Crematorium, demand, district heating, heat, investment, network, NPV,

(3)

PREFACE

This is a Master thesis work on Hovdestalund crematorium which is a mission from

Mälarenergi. The author of the report is Charlie Lindquist from Mälardalens University in the master’s program of sustainable energy systems.

I would like to give my thanks to Mälarenergi for giving me this opportunity to do my thesis with the help of you. I would also like to give a special thanks to my supervisor Jesper Ericson at Mälarenergi for all the help I have received.

Västerås in June 2021

(4)

CONTENT

1 INTRODUCTION ...1 1.1 Background ... 1 1.2 Purpose/Aim ... 2 1.3 Research questions ... 3 1.4 Delimitation ... 3 2 METHOD ...3 3 LITERATURE STUDY ...4 3.1 Technical options ... 5

3.1.1 District heating unit. ... 5

3.1.2 Cooling tower. ... 6

3.2 District heat connection possibilities ... 7

3.3 Energy investigation Hovdestalund ... 7

3.4 Industrial waste heat potential ... 9

3.5 Solar heating connection to grid ...10

3.6 Economic analysis of a flue gas condenser ...11

3.7 Economic analysis of a gas-fired absorption heat pump ...12

3.8 District heating pricing ...12

4 CURRENT STUDY ... 14

4.1 Pipeline and heat demand ...15

4.2 Connection solutions ...16

4.2.1 Technical solution 1 ...16

4.2.2 Technical solution 2 ...17

4.3 Heat exchanger design ...18

4.4 Calculations ...18

4.4.1 Case 1: Kadesjös 400kWh ...19

(5)

4.5 LCA ...20

5 RESULTS ... 21

6 DISCUSSION... 28

CONCLUSIONS ... 31

7 SUGGESTIONS FOR FURTHER WORK ... 31

(6)

LIST OF FIGURES

Figure 1 DHU with primary and secondary circuit ... 5

Figure 2 Differential pressure curve ... 6

Figure 3 Solar panels connection, principal sketch. ... 10

Figure 4 Augustenborg connection sketch. ... 11

Figure 5 Filter system ...14

Figure 6 New facility and pipe placement ... 15

Figure 7 Technical solution 1 ... 17

Figure 8 Technical solution 2. ... 17

Figure 9 Heat exchanger temperatures. ... 18

Figure 10 Result: Technical solution 2 ... 22

Figure 11 Sold/bought heat: 400kWh per cremation. ... 23

Figure 12 Self-consumption: 400kWh per cremation. ... 23

Figure 13 Sold/bought heat: 242,5kWh per cremation. ... 24

Figure 14 Self-consumption: 242,5kWh per cremation. ... 24

Figure 15 Demand & heat recovery. ... 25

Figure 16 Demand, heat recovery & bought heat. ... 25

Figure 17 NPV Case 1 ... 27

Figure 18 NPV Case 2 ... 27

LIST OF TABLES

Table 1 Water requirements for different types of plants ... 7

Table 2 Kadesjös input variables. ... 8

Table 3 Kadesjös LCC input variables. ... 8

Table 4 Economy ... 11

Table 5 Stockholm exergi pricing ... 13

Table 6 Mälarenergi pricing ... 13

Table 7 Vattenfall pricing ... 13

Table 8 Normal year correction ...16

Table 9 Pipeline cost ... 20

Table 10 Pricing, sensitivity analysis...21

Table 11 400kWh heat per cremation across one year... 22

Table 12 242,5kWh heat per cremation across one year... 22

(7)

NOMENCLATURE

Symbol Description Unit

C Investment cost SEK

P Power W

r Discount rate %

T Time years

ABBREVIATIONS

Abbreviation Description

AHP Absorption heat pump CHP Combined heat and power DH District heating

DHU District heating unit FGD Flue gas desulfurization FVB Fjärrvärmebyrån HX Heat exchanger

IRR Internal rate of return LCA Life cycle analysis ME Mälarenergi NPV Net present value PUR Polyurethane

(8)

DEFINITIONS

Definition Description

Block Boiler combined with a turbine

(9)

1

INTRODUCTION

Waste heat might be considered a free source of energy, by investing in solutions that can use the heat more efficiently would seem economically feasible at first glance. This is not always the case due to high investment costs and potential risks such as economic uncertainty. Crematorium ovens are operating at a very high temperature without utilizing possible excess heat for district heating as an example. Using the energy created from the crematorium Berthåga could save roughly 1500 MWh which translates to heating up hundreds of buildings (Harnesk, 2015). During the 40 latest years, the heat market has been growing due to

increasing amount of heating surfaces. The direction of the increase is showing that the increase is becoming less and eventually leading to a reduced heat demand. Having access to a safe and reliable heat supply is an important part of society and today’s market now faces several challenges. Energy efficiency goals, competition between different heating solutions, and politics are some of these challenges. The Swedish heat market sells 100TWh and is trading 100 billion SEK annually. Small houses accounts for more than half of the trading price and roughly 40% of the heat demand, apartment buildings stand for 30% and premises for 25% of the demand. industrials are the smallest consumer group. The district heating market accounts for roughly half of the total heat demand while electrical heating and heat pumps together accounts for one third. (Sköldberg & Rydén, 2014)

The reasoning why this work has become relevant today is because the Swedish church will expand Hovdestalund with additional ovens. By building a new facility will create the option to consider energy efficiency when looking into the technical aspects. Utilizing possible energy production from a crematorium facility is considered ethically wrong by some people and there for fear of negative reactions has delayed the question for some crematoriums. In an article from the website “the telegraph” Joerg Michner writes about the use of recovering energy from the crematorium in Halmstad. The director of the cemetery Lennart Andersson had the plan to use the waste energy and created a plan towards the goal and was met with approval from locals. Andersson says, “Of course it’s possible that there will be some discussion about the ethics of this, but from our side, this is a purely environmental idea. There will be no difference in the ashes”. (Michner, 2008)

1.1 Background

This study will contribute with a price model concerning Hovdestalund crematorium, in the creation of the model several aspects must be considered. To have a sustainable relationship between ME, Hovdestalund or other future potential customers, a fair price must be achieved to satisfy and benefit all sides involved. Choosing the correct connection solution will play a crucial part in the results as both parties are affected. Hovdestalund will benefit from a design that allows for high recovery potential and ME will benefit from a solution that

(10)

simplifies billing, maintenance and similar. A heat recovery system can benefit by using more of the input energy and a higher efficiency of the heat source can be thus be achieved. Using some sort of heat exchanger, the outlet temperature of the flue gas can be lowered and thus leading to the higher efficiency. Lower outlet temperature also means lesser emissions. A study made by Priedniece et al., (2018) shows that using a condenser on a pellet boiler could reduce the outlet flue gas temperature from a span of 126-132˚C down to a span of 21-48˚C. This showed an increase in efficiency of the boiler between 5-15%. (Priedniece, et al., 2018) The possible heat recovery from flue gas cooling from a single cremation lies roughly between 230-255 kWh (Norbeck, Karolina, 2018); (SKKF, 2018). SKKF (2018) also states that there are 23 crematoriums today in Sweden that recovers heat to the district heating network and that the temperature is often higher than 90-95˚C. Lidén (2006) writes about the new crematorium in Kramfors that was going to be built in 2008, the crematorium would be contributing to district heating. With flue gases of 1100˚C, a heat exchanger is needed to reduce the gases down to 120˚C. The purpose is to install a purification of the flue gases due to high emissions of mercury. Anders states that crematoriums have become some of the biggest sources of mercury emissions where they release approximately 0,2-0,5 grams per cremation. With the number of cremations per year, the emissions will lead to several hectogram per year. In theory, only a few grams are needed to poison a lake. (Lidén, 2006) In Västerås, all power- and heat production comes from Mälarenergi CHP plant and different blocks. The CHP plant started being built in 1960 and has since been expanded with several boiler blocks. The production reaches 700 GWh electricity and 1800 GWh heat per year. (Mälarenergi, 2021b) Mälarenergi’s newest boiler is block 7 which uses wood waste from households and industries as fuel. Block 7 has an installed capacity of 150 MWth (thermal effect) and accounts for approximately 36% of the heat production (Mälarenergi, 2021a). Block 6 is the biggest producer of heat which accounts for 50% of the total heat production, the boiler was set to drift in 2014. Block 6 uses both waste and biofuels as fuel and has an installed capacity of 167 MW. The reasoning for using waste in the fuel mix is to reduce the use of peat and thus reducing the emissions of fossil carbon dioxide with 300 000 ton per year. The boiler burns roughly 60 tons of waste per hour. (Mälarenergi, 2021c) Together, block 6 and 7 stands for 86% of the heat production for Västerås, Surahammar,

Hallstahammar, Skultuna, and Kolbäck (Mälarenergi, 2021a). In Västerås, 98% of all estates are connected to the district heating network and the total length of the network in 860 km. The district heating network is a closed system that delivers heat from a central facility in the form of warm water. The water is heated up to 65-110 degrees Celsius and then pumped out into the network. (Mälarenergi, 2021d)

1.2 Purpose/Aim

The aim of this study is to investigate possibilities of recovering otherwise wasted heat from Hovdestalund crematorium by condensation of the flue gases. The study will investigate technical and economic possibilities of connecting the facility into the district heating network where the facility can be both a consumer and producer.

(11)

This is done as a mission from Mälarenergi to get an insight in the costumer’s alternative cost and creating a business model to deliver waste heat into the district heating network.

1.3 Research questions

- What technical solution should be used for the costumer to be connected as producer and consumer to the grid?

- How will the costumer’s production and consumption pattern of district heating look like across the full year?

- What is the long-term perspective and business perspective?

- How will the energy from the crematorium affect Mälarenergi’s grid and other production?

1.4 Delimitation

This study will be directly focused on Mälarenergi’s district heating network and Hovdestalund crematorium facility. The foundation of the report can be used for other district heating networks or crematoriums, but the values, parameters, and underlay will directly correspond to the studied areas. The study will only regard the step after the flue gases have been cooled down. The cooling step will be some form of heat exchanger such as a condenser. The condenser will have the same dimensions no matter which solution is picked. Therefore, this step will not be included into the analysis as it would be the same for both cases. The study will be choosing between two solutions. First, a cooling tower solution which lowers the outlet temperature without using it for district heating. Secondly, a district heating unit which will use the flue gases to heat up returning district heating water before supplying it into the pumping hot water.

2

METHOD

A literature review will be conducted to acquire information previously made for similar topics and technical aspects. The literature will give pre knowledge about connecting systems into the district heating network, techno economic and environmental impact. Technical solutions will give the foundation of the project where the selection of heat recovery method will be determined. With technical solution decided, corresponding calculations can be made to design the system. With Excel, calculations, tables, and figures can be made for each case. With the help of the IF-function and the previous year’s heat demand, the heat will be

determined if it will be used for self-consumption, sold to the grid, or bought from the grid. A life cycle analysis (LCA) will also be made using the net present value (NPV). The LCA will be based on market heat price, if the recovered heat is sold, bought, or self-consumed, laying the new pipeline and more. Data from the crematorium will be received from interviews of

(12)

workers from Hovdestalund and other personal of the Swedish church. Mälarenergi (ME) will assist with heat consumption data of the already installed facility from previous years which will be used when comparing each case. The consumption data will be reworked so that it is representing a normal year when it comes to weather conditions. After the design is complete, an economic evaluation can be made for the systems from both the customers and ME’s perspective. With help from ME’s previous knowledge and competence of district heating units and cooling towers, prices of installation, assembly and more can be set for the different solution designs.

3

LITERATURE STUDY

Heat recovery will become more viable with new environmental goals such as CO2 reduction.

Heat recovery methods are used to reduce emissions, improve efficiency and to save

otherwise wasted energy. Waste heat means heat that is bound to either liquids or gases that are emitted from processes to the surrounding and that cannot be used. Waste heats have two sub-concepts, primary- and secondary waste heat. With primary waste heat (high-grade heat), the temperature is high enough to be used directly in the district heating (DH)

network. Secondary waste heat (low-grade heat) is heat that is too low to be used directly, the heat must therefore be pre-heated by a heat pump. (Cronholm, Grönkvist, & Saxe, 2009) Using a heat pump will increase the temperature to a more viable level, the drawback with a heat pump is that it will use more electricity to power the pump and depending on case and time of year, this could be sub-optimal since the grid will be affected. Therefor it is better to take advantage of high-grade heat where no other form of energy conversion will take place, the high-grade heat can be used directly for DH.

A boiler that is using a preheater can reduce the flue gas outlet temperature to between 100-150˚C which is a lot higher than the average dew point temperature of the flue gas which is around 55-60˚C. Because the flue gas is not cooled down to the dew point temperature, the flue gas only releases part of the sensible heat and most of the latent heat is not recycled. A self-driven wet-hot flue gas total heat recovery system was studied by Zhang, Yang, Fan, and Huang. With the flue gas outlet temperature reduced to 41,39˚C, and with a relative humidity of 61,25%, the maximum heat recovery rate could reach 11,6%. Increasing the flue gas inlet temperature from 350-550˚C, the heat exchange capacity of the system increased by 46%. (Zhang, Yang, Fan, & Huang, 2020) The cooling water to flue gas flow rate ratio will also strongly affect the rates of heat transfer and condensation and condensation efficiency (Levy, Bilirgen, & DuPont, 2011).

(13)

3.1 Technical options

Hellborg (2017) states that a heat exchanger (HX) will transfer heat from one source to another, a commonly used HX is the shell and tube model or the plate model. Typical heat exchangers are condensers, evaporators, and radiators where the mass flow can either run co-flow, crossflow, counter-flow, or a combined variant. In a shell and tube HX, one fluid runs in the inside of the tubes and the other fluid ruins in the shell that is surrounding the tubes. Hellborg (2017) also states that “the geometry allows for high pressure and flexibility concerning phases as well as the possibility to use finned tubes for increased heat transfer” (p. 14).

3.1.1 District heating unit.

A district heating unit (DHU) is used at every house or facility that is connected in the DH network. The DHU construction contains a heat exchanger, control equipment, safety equipment and more. The DHU can be constructed based on the purpose but most

commonly, the unit is prefabricated. The DHU will separate the primary DH water circuit from the costumer’s secondary water circuit. The reason for this is due to the differences in pressure between the two circuits, the primary circuit will have a higher pressure to pump around the water across the network. The temperature level will also differ between the two circuits.

Figure 1 DHU with primary and secondary circuit

The F:101 is a provision which is a technical standard for Swedish district heating units. F:101 tells how a building should be adjusted to a DH system, what requirements apply regarding dimensioning, execution, installation, operation, and maintenance. According to F:101, the conventional primary system will have a pressure of 1,6MPa and a supply temperature of ≤100˚C. A low temperature system will have the same pressure but with a lower temperature of ≤80˚C instead. A secondary system will have a differential pressure of 0,6-1,0MPa and a

(14)

supply temperature of ≤80˚C. The system will have differential pressure between

0,1-0,6MPa, the amount will depend on the distance from the distribution central. The closer the DHU is to the central, the higher the differential pressure is. The pump of the distribution system will base the flow on the DHU with the lowest differential pressure, it cannot be lower than 0,1MPa. The differential pressure is illustrated in Figure 2 where A, B, C, and D are DHU’s. (Svensk fjärrvärme , 2014)

Figure 2 Differential pressure curve

3.1.2 Cooling tower.

Apana (2021) states that cooling towers can be used to remove excess heat from different systems or equipment by either water evaporation or airflow. Cooling towers are commonly used for power plants, factories, refineries, and similar systems that depends on water circulation to reject heat. Cooling towers share the same process of where hot water enters, and cold water exits. The most common cooling tower is the open circuit (wet cooling) cooling tower where the hot water is drained or sprayed into a spill area. In the spill area, some of the water evaporates and the rest collects in the basin. From air circulation, the water that does not evaporate will cool down and then be recirculated back into the system. Another type of cooling tower is the closed circuit, here the water runs in a tube that gets sprayed with cooling water and a fan that will help with heat transfer.

There are many types of cooling tower types but what most have in common is that they use a lot of water in the process. The amount of water that will be used is based on the size of the cooling tower. Due to the high usage of water, cooling towers have high upkeep costs. Some distribution systems such as forced draft and pressurized water require more power but all of them require regular maintenance (Apana, 2021). World nuclear association (2020) has written an article about cooling power plants and water requirements from different types of plants were calculated. From Table 1 the amount of water for each type can be seen. Coal plants are assumed to need flue gas desulfurization (FGD), this will increase the water usage. (World nuclear association, 2020)

(15)

Table 1 Water requirements for different types of plants (World nuclear association, 2020) Coal, once-through, subcritical, wet FGD 0.52 liters/kWh

Coal, once-through, supercritical, wet FGD 0.47 liters/kWh Nuclear, once-through, subcritical 0.52 liters/kWh Coal, recirculating, subcritical, wet FGD 1.75 liters/kWh Coal, recirculating, supercritical, wet FGD 1.96 liters/kWh Nuclear, recirculating, subcritical 2.36 liters/kWh

3.2 District heat connection possibilities

The crematorium will be used as a prosumer of heat, the concept of prosumer means that the heat can be supplied to the DH network as well as extracted from the network. Being

connected as a prosumer can be done in two ways, either by series connection or by parallel connection. With series connection, the produced heat is stored in a heat storage. If the heat storage temperature becomes too high, some of the heat will be supplied into the DH

network. With this type of connection, the input power does not need to match the power from the locally produced power. With parallel connection, the local heat can either be used in a heat storage or fed into the network (Lennermo, Lauenburg, & Brange, Små värmekällor, 2016). A heat source can be connected into the distribution network by primary connection in four different ways according to Lennermo & Lauenburg (2016).

- Return/supply (R/S): In a R/S system the water is taken from the return pipe (primary side), heated up to desired temperature and then fed back into the supply pipe.

- Return/return (R/R): In a R/R system the water is taken from the return pipe, heated up to any temperature and then fed back into the supply pipe.

- Supply/return (S/R): In a S/R system the water is taken from the supply pipe, heated up to any temperature and fed back into the return pipe.

- Supply/supply (S/S): In a S/S system the water is taken from the supply pipe, heated up to any temperature and then fed back into the supply pipe.

S/R and S/S are uncommonly used as they feed the DH networks return line with higher temperature or heats up the supply line more than necessary. The R/S system is the most beneficial system which does not affect the return temperature, but it will reduce the heat power load from the central distribution unit. (Gunnar Lennermo; Patrick Lauenburg, 2016)

3.3 Energy investigation Hovdestalund

An energy investigation on Hovdestalund has been made by Kadesjös (2020) which compared two alternative cases for the newly built facility. Case 1 of Kadesjös report is the possibility to connect to the district heating grid, and case 2 is looking at geothermal as an option. Both cases will cover the energy need for the existing crematorium, economy building

(16)

and the new crematorium. The study is comparing the two cases with pros and cons and an overall LCC analysis for both options. The study has calculated the heat demand and

assumed that the recycled heat will cover 50% of the yearly energy use. In Table 2, the input variables that Kadesjös has used is shown.

Table 2 Kadesjös input variables.

Input heat data

Total assumed oven power 300 kW

Total assumed heat demand 900 MWh/year Power per cremation 400 kWh

Cremations 6 Daily

Operation 200 Days per year Cremations 1200 Yearly

Possible energy recovery 480 000 kWh/year, oven

Input data cooling

Total assumed cooling power 80 kW

Total assumed energy 15 MWh/year Assumed COP 3 One part

electricity gives 3 parts cooling

In Table 3, the LCC input variables that Kadesjös has used is shown. Table 3 Kadesjös LCC input variables.

Input data LCC

Lifespan 20 years

Interest rate real 3 % Maintenance cost increase 2 %

Heat cost 1 SEK/kWh

Heat price increase 2 % Cooling cost 1,2 SEK/kWh Cooling price increase 2 % Electricity price 1,2 SEK/kWh Electricity price increase 2 %

The result of Kadesjös report shows that the investment cost is higher for case 2, 4M SEK compared to case 1 of 1,5M SEK. The yearly cost of bought energy is higher for case 1 with nearly 8,229M SEK compared to 3,248M SEK for case 2. The maintenance cost for case 1 is lower than case 2, 271t SEK and 1,552M SEK, respectively. After the LCC analysis the final cost came to a total of 9 999 455 SEK for case 1, and 8 800 117 SEK for case 2. The conclusion of the investigation is that across the expected lifespan of 20 years, case 2 will save 1M SEK compared to case 1. With geothermal, 150 MWh are bought yearly compared to 450 MWh for DH. (Kadesjös, 2020)

(17)

3.4 Industrial waste heat potential

Ziemele, Kalnins, Vigants, Vigants, & Veidenbergs (2018) made a study about waste heat potential and its integration in a fourth-generation district heating system. The waste heat of the factory will come from sources such as wastewater, smoking vents, cooker exhaust, flue gases from boilers and heat from condensers. The waste heat will be used for preheating air and water supply or for transfer to DH network. The authors created an algorithm of methodology that they worked from. The first step, initial data collection, determination of assumptions and evaluation of waste heat sources. The second step is to analyze the

individual variables and thereafter create a calculation on the potential of waste heat. With the results achieved, different heat recovery scenarios are created and then an economic feasibility analysis to know whether the result is acceptable or not. The three main waste heat sources the study focused on are:

- Flue gases from boilers - 800˚C. - Hot wastewater - ~95˚C.

- Hot air from production facilities - ~35-40˚C.

An indirect flue gas condenser was chosen to recover the heat from the flue gas. To calculate the possible power of the condenser Ziemele et al. (2018) used the equation below.

𝑃𝑘 = 𝐿𝑑𝑔∙ (𝐻1− 𝐻2)

To calculate the theoretical heat recovery power of the heat exchanger of the wastewater they used the equation below.

𝑃𝑤 = 𝑚 ∙ 𝐶𝑝∙ (𝑡1− 𝑡2)

Since the factory is not connected to the DH network, new pipes are needed for connection to the system. To calculate the flow and the optimal pipeline diameter they used the two

following equations. G is the flowrate (kg/s), and d is the diameter. 𝐺 = 𝑁𝑚𝑎𝑥

𝐶𝑝∙ ∆𝑡

𝑑𝑖𝑛𝑛𝑒𝑟= √

4 ∙ 𝐺 𝜌 ∙ 𝜋 ∙ ѵ

For economic evaluation, the NPV and Internal rate of return (IRR) was used. 660 kW was recovered from using the flue gas condenser and heat exchanger from hot wastewater. The recovered energy equals 10,4% of the primary energy input of the factory. A total heat recovery of 93,2 MWh monthly. After considering the new-built pipeline of 2000m with an inner diameter of 82,5mm, heat losses resulted in a decrease of heat down to 67,5 MWh. They state that depending on the season, 4-12% of the heat load monthly can be substituted using the recovered waste heat. This will in its turn reduce CO2 emissions by 106 tons per

season, a reduction of 5-10% monthly. Using the least profitable energy tariff of 126€/MWh instead of the average 40€/MWh, the payback period of the investment is 10,4 years. (Ziemele, Kalnins, Vigants, Vigants, & Veidenbergs, 2018)

(18)

3.5 Solar heating connection to grid

A report by Dalenbäck, Lennermo, Andersson-Jessen & Kovacs (2013) has investigated and evaluated the integration of 22 already existing solar power plants into the district heating networks. The investigation has focused on points such as system design, installation,

operational follow-up, and thermal yield. With the information from this study, Dalenbäck et al (2013) states that this can be the foundation for newly built facilities. A principal sketch from the study shows how the solar panels are connected to the DH network, this can be seen in Figure 3Error! Reference source not found.. In the figure, there are two district heating units (DHU), one that is connected to the solar panels (far right rectangular), the other one is for internal use such as tap water and radiators.

Figure 3 Solar panels connection, principal sketch. Inspired by: (Dalenbäck, Lennermo, Andersson-Jessen, & Kovacs, 2013)

The 22 solar collector systems have basically all the same system structure, one solar collector circuit with temperature sensors, a pump, a heat exchanger, and a district heating circuit with temperature sensors, a pump and one or more valves. From this system

structure, the plan is to deliver heat with the same constant temperature into the DH

network. To deliver the constant temperature, the three-way valve (SV31) will open when the required temperature is reached. The systems connection sketch that is used for ex.

(19)

Figure 4 Augustenborg connection sketch. Inspired by: (Dalenbäck, Lennermo, Andersson-Jessen, & Kovacs, 2013)

The conclusion that was made was that all facilities are working but there are a lot of flaws coming from the different stages in the project. There are flaws in the planning face, in their behavior, in operation, and maintenance, this has overall affected the overall function. (Dalenbäck, Lennermo, Andersson-Jessen, & Kovacs, 2013)

3.6 Economic analysis of a flue gas condenser

The study by Terhan & Comakli (2016) investigates the event of condensing water vapor flue gas to recover latent heat from a 60MW gas fired district heating system of a university. They state that the condensing and non-condensing part should be examined separately because the classical heat exchanger design method becomes deficient. The total investment cost comes to a total of $ 83,711.16 where the annual savings in fuel is $ 407,396.16. They state that the payback period is less than 1 year.

Table 4 Economy (Terhan & Comakli, 2016) Economy

Economic life 20 years

Interest rate 5%

Working hours 19

Yearly working hours 6935 Electricity cost $ 0,12/kWh Tube material $ 8,78/m

With their flue gas condenser design, they manage to decrease the exit flue gas temperature from 150˚C to 40˚C. The heat exchanger condensation efficiency is 17,8%. (Terhan & Comakli, 2016)

(20)

3.7 Economic analysis of a gas-fired absorption heat pump

Lu et al. (2019) has made a study of a novel gas-fired absorption heat pump (AHP) that is used for district heating. The study will see the possibility to recover high-grade sensible heat from the preheater and low-grade latent heat from the intermediate evaporator of the flue gas. The study shows that the system can cool the flue gases from 180˚C down to 20˚C, from this 5,8kW waste heat is recovered. From the 5,8kW, 2kW is recovered from sensible heat and 3,8kW is recovered from latent heat. The differences between the proposed system and a single-effect system are that the proposed system uses an intermediate evaporator, an

intermediate absorber, a preheater, and a separator instead of a condensing heat exchanger. The conclusion of the study shows that the system will save 11,7% and 39,6% of primary energy when compared to the single-effect AHP and the gas-fired boiler. The proposed system with waste heat recovery and the gas fired boiler will have a payback time of 2,9 and 2,5 years, respectively. (Lu, et al., 2019)

3.8 District heating pricing

Creating a reasonable pricing model will be important for both ME’s and Hovdestalund’s perspective, it should favor both sides as equally. ME can buy the heat but compared to their alternative cost of production, it will always be more expensive. To create a model, similar businesses such as Stockholm Exergi and Vattenfall has been researched as a validation towards creating this model. Comparing these companies price policies with ME’s already existing one will set as a baseline for Hovdestalund’s recovery price model and possible future heat recovery models.

Stockholm Exergi is the main supplier of DH in Stockholm. Stockholm exergy has 800 000 costumers. The DH network consist of 86% renewable and recyclable energy and a network length of a 3000 km. (Stockholm exergi, 2021a) Stockholm exergy are setting their prices based on five principals that are listed below.

• Option pricing: The prices should be competitive with other options such as geothermal heating.

• Price stability: A stable, long term, predictable price development.

• Equal treatment: Equivalent costumers who buys similar products will have an equivalent price.

• Freedom of choice: Customers might have different preferences about the price structure, there are therefore some optional choices.

• Openness: Easy comparison between DH and other options.

Stockholm exergy base their prices on three components, power, energy, and return

temperature bonus or -fee. Power is divided into two parts, power price (SEK/kW) and power tax (SEK/year). The total average cost per kW is lower for higher power levels, the reason for this is to reflect the economies in an alternative heating solution. The energy price is divided into two periods, during April-October, there will be a lower price. During November-March,

(21)

of the water, at a lower temperature than 50˚C, there will be a bonus while having a temperature above will result in a fee. The bonus/fee will only take place in the period November-March. (Stockholm exergi, 2021b)

Table 5 Stockholm exergi pricing (Stockholm exergi, 2021b) Power

Power level [kW] Power fee [SEK/year] Power price [SEK/kW]

0-99 0 876 100-499 2 600 850 500-999 74 600 706 1000-2499 175 600 605 2500 and above 370 600 527 Energy

Period Energy price [SEK/MWh]

Apr-Oct 250

Nov-Mar 656

Return temperature [Nov-Mar]

Temperature [˚C] Bonus [SEK/MWh,˚C] Fee [SEK/MWh,˚C]

Below 50˚C 6,3 -

Above 50˚C - 20,5

ME has a similar way of price charging for DH, a high price during winter period, slightly lower during spring and autumn, and a cheap price during summer when the demand is lower. The DH price is shown in Table 6. (Mälarenergi, 2021f)

Table 6 Mälarenergi pricing (Mälarenergi, 2021f)

DH price Mälarenergi

Jan-Feb, Dec 680 [SEK/MWh] Mar-May, Sep-Nov 630 [SEK/MWh] Jun-Aug 220 [SEK/MWh]

Vattenfall operates in several locations in Sweden, with nine DH plants, Vattenfall provides heat and warmwater to hundreds of thousands of buildings and companies. (Vattenfall, 2021a) Vattenfall will charge two different pricing, one during summer, and one during winter, the price is taken from Uppsala commune. The DH price is shown in Table 7. (Vattenfall, 2021c)

Table 7 Vattenfall pricing (Vattenfall, 2021c)

DH price Vattenfall

Summer 103 [SEK/MWh] Winter 835,25 [SEK/MWh]

Vattenfall charges 45 000 SEK for complete connection and installation of DH excluding heat exchanger. This price is only up to 15-meter connection length, above this will increase the

(22)

price with 4375 SEK/extra m. (Vattenfall, 2021b) This pricing is set for private costumers, which means this could differ slightly for companies.

4

CURRENT STUDY

An interview with Johansson (2021) from Hovdestalund gave information about the cremation process. Today Hovdestalund has two ovens operating, the Swedish church are planning to expand the facility with two new ovens. With this new expansion, the Swedish church is looking for alternative ways of saving energy. In general, roughly 90% of people will be cremated and the rest will be buried. Most burials are because of religious reasons. In 2020, approximately 2300 cremations have been made where a cremation will take somewhere around 70-80 minutes. Previous year, there was about 2100 cremations. The increase was likely caused by the pandemic. (Johansson, 2021) The current system of

Hovdestalund is illustrated in Figure 5Error! Reference source not found.. The exhaust flue gas from the cremator is led into a flue gas cooler where the gases are cooled down from abought 600˚C down to about 120˚C. Water is pumped around and is cooled by an external cooling circuit to be used again in the system. In the pre-separator, particles are separated from the gases before fed into a filter unit. In the filter unit, mercury, and other dangerous substances are removed from the gases. Once the gases have been purified, they are fed into an exhaust fan that pushes it out of a chimney.

(23)

4.1 Pipeline and heat demand

Hovdestalund is split in two separate buildings, one crematorium and one economy building which is shown in Figure 6. Beside the economy building, the new crematorium will be built. The exact location is not certain yet, but this is a rough estimation from the three possible locations that is mentioned in a report by the Swedish church. Change in location will only differ with a couple of ten meters when considering DH pipe length. Hovdestalund is

currently connected to the DH network where the blue line starts in Figure 6. The new facility is represented by the bright red box and the blue line represents the intended way to place the pipes. The length of the new pipeline is around 140m, with roughly 60m asphalt and 80m green space.

Figure 6 New facility and pipe placement

Using recorded heat data for all Hovdestalund’s buildings will require a change in values so that they will represent a “normal year”. 2020 was considered a hot year overall, by

correcting the data, the values will represent average weather conditions. This is done by taking the normal year constants for each month and multiplying them with the data for each recorded hour. With the data for the three past years from 2018-2020, heat data will be presented and calculated on an hourly basis with a normal year correction. After each year has been re-calculated to represent a normal year, all three years show a similar heat demand. Therefore, the year 2020 will be used as a reference for further economic calculations. The correction constant for each month and year can be seen in

(24)
(25)

Table 8 Normal year correction (Mälarenergi, 2021e)

Normal year correction

2018 2019 2020 JAN 1,108 1,004 1,364 FEB 0,993 1,213 1,272 MAR 0,956 1,125 1,178 APR 1,147 1,146 1,127 MAY 1,897 0,974 0,877 JUN 1,235 1,241 1,371 JUL 1,636 0,828 0,796 AUG 1,198 1,07 1,185 SEP 1,205 1,012 1,121 OCT 1,124 0,971 1,106 NOV 1,083 0,993 1,212 DEC 1,037 1,106 1,183

4.2 Connection solutions

The following two solutions are considered in this study. When deciding between the two, the decision will be based on how well it will benefit both ME and Hovdestalund.

4.2.1 Technical solution 1

Solution 1 is based on the report by Dalenbäck et al. on solar heating in DH systems that has a similar connection between the primary and secondary system. In this configuration, there will be three circuits, the primary DH circuit, the secondary oven circuit, and another

secondary circuit for own use. The district heating network will supply the internal-DHU with heat when the ovens are not up and running. Once they are running, the oven-DHU will supply the DH network with heat instead.

(26)

Figure 7 Technical solution 1

4.2.2 Technical solution 2

Solution 2 will pair the oven circuit and own use circuit together, creating one single circuit. A control unit will tell a three-way valve if the own use circuit has a high enough temperature, if not, the valve will open and feed heat from the oven circuit until it reaches the desired temperature. If the ovens are supplying heat, the valve on the primary circuit will close so that no heat transfer comes from the DH network. Like the first solution, if there is no need for own use, the heat will be transferred and sold into the DH network.

(27)

4.3 Heat exchanger design

To achieve a price suggestion of the DHU, the HX needs to be designed based on temperature and power output. The HX will have a power output of 600 kW which corresponds to two working ovens. The temperature needs to be estimated to determine all four sides of the HX. The temperature achieved after the ovens and the temperature that is being fed back into them is received from an interview with Patrik Selinder from Fjärrvärmebyrån (FVB). According to Patrik, the temperature than can be expected coming out of the ovens are somewhere around 80-95˚C based on season. The temperature going back into the ovens are 2-3 degrees higher than the return temperature of the primary network. (Selinder, 2021) Based on previous data from ME, the expected return temperature in summer lays the baseline of the design. The return temperature is between 46-49 degrees. Meaning the temperature going back into the ovens are 48-52 degrees. The last temperature is the outlet temperature into the primary system from the HX, the expected temperature from a well-designed HX is assumed to lose around 3 degrees from the heat transfer. This results in an outlet temperature of around 77-92 degrees that is fed into the DH network. To design the HX, temperatures need to be set without a span, all the temperatures chosen are illustrated in Figure 9.

Figure 9 Heat exchanger temperatures.

4.4 Calculations

Looking at the heat demand data received from ME for Hovdestalund, heat sold, bought, and used for self-consumption are calculated in Excel. Adding operating hours and days will be the foundation for retrieving these values. Operating hours will be set to 08:00-16:00 where they only run during weekdays, and a total of 200 operating days. From this, an IF-statement is created which will calculate if the facility needs to buy, self-consume, or sell excess heat generated by the system. The IF-statement will check if the hours and days are within

(28)

operation time and then see if the hourly demand from 2020 is higher or lower than the recovered heat. If the demand is higher, the facility will use all recovered heat and buy the rest from the grid. If the demand is lower, the facility will instead sell the excess heat to the grid after using what is needed to cover the demand.

The three following sub-sections will consist of three cases of which this report will base its results on. Case 1 will follow Kadesjös investigation and choose their heat recovery potential of 400kWh per cremation. Case 2 will follow the literature study research and choose the found heat recovery potential of 242,5kWh per cremation. Lastly a short cooling tower investigation is looked in to since it is what the facility is currently using in their process.

4.4.1 Case 1: Kadesjös 400kWh

To calculate the hourly heat recovered from the ovens, the number of cremations per year is divided with the operation days. This results in roughly 12 cremations per day, 6 cremations per oven. 12 multiplied with the recovering potential (400kWh for case 1) results in a heat recovery potential of 4,8MWh per operation day. By taking the daily recovery potential of 4,8MWh and dividing that with eight operating hours, a value of 0,6MWh heat per hour will be recovered from the process while the ovens are running. If the hourly heat demand is 0,3MWh during operation, 0,3MWh will be used for self-consumption and the remaining 0,3MWh will be sold to the grid.

The heat demand of 2020 was 852,3MWh heat, and from the calculations, the recovery system sold 750MWh, bought 643,1MWh and self-consumed 209,2MWh.

By talking with Einar Port from ME about how much they are willing to buy heat for from the crematorium, the conclusion wasa price changing based on season. ME has a fuel cost at the CHP (combined heat and power) plant for each month, from these costs, the crematorium can sell their heat for that specific monthly price. When buying heat from ME, the

crematorium will buy for ME’s listed price in Table 6. The recovery system can then sell heat for 90 546 SEK, buy heat for 414 623 SEK, and save 135 784 SEK by self-consuming instead of buying.

4.4.2 Case 2: Literature study 242,5kWh

In the literature study that was made for this study, two reports that were found says that the heat that can be recovered per cremations lies between 230-255kWh. This will result in a daily heat recovery of 2,91MWh, and an hourly heat recovery of 0,364MWh compared to 0,6MWh for Kadesjös. This results in less heat sold while self-consumption and bought heat remains the same.

With the same heat demand of 852,3MWh, the recovery system sold 375,8MWh, bought 643,1 MWh and self-consumed 209,2MWh. With the same seasonal costs, the recovery system can sell heat for 44 972 SEK, buy heat for 414 623 SEK, and save 135 784 SEK by self-consuming instead of buying.

(29)

4.4.3 Cooling tower

A cooling tower will not be able to recovery any heat for either self-consumption or selling excess heat to the grid. The cooling tower is only needed to cool down the flue gases to reduce emissions. Choosing the cooling tower solution will result in the crematorium having to buy their full heat demand from the DH grid. By using the seasonal values from ME Table 6, this results in the facility having to buy heat for 550 407 SEK per year to cover up the demand. This price only comes from variable energy costs and does not consider additional power-, investment-, and maintenance costs.

4.5 LCA

With the heat demand, yearly costs of bought, sold, and self-consumed heat can be used to calculate the NPV. According to Mälarenergi (2021d), the lifetime of a DHU is arounds 25 years where some smaller parts might need a change before that. Therefor the LCA analysis will be made for around 25 years whereas after that time, the DHU will probably need to be replaced and thus need some reinvesting. To investigate how the long-term economic value will unfold, the NPV will be calculated for each case, see equation 1 below. C is the investment cost, r is the discount rate and T is the time in years. The discount rate is assumed as 6% which is taken from TheLocal (2020) and Trafikverket (2014). Cash flow will in this case be the heat sold to the district heating system along with the saved cost from self-consumption. The investment cost will contain the pipeline cost and the DHU cost.

𝑁𝑃𝑉 = −𝐶 + ∑ 𝐶𝑎𝑠ℎ 𝑓𝑙𝑜𝑤 (1 + 𝑟)𝑖 𝑇

𝑖=1

𝐸𝑞. 1

With the help from experts from ME, a rough estimate price on laying down a new pipeline is achieved. Laying the pipeline will result in costs of 8600 SEK/m asphalt, and 6400 SEK/m green space, this pricing does not consider other terrain such as rock. This will result in a total cost of 1 028 000 SEK, values for the cost in shown in Table 9. The new pipeline will be a PUR (Polyurethane) pipe, a PUR pipe consists of two smaller pipes, one pipe for supply water and the other for return water. The size of the outer pipe is 250mm in diameter, and 80mm diameter on the inner pipes. The PUR pipe is isolated with PUR foam, hence the name.

Table 9 Pipeline cost

Pipeline 140m

Cost green space 6400 SEK/m Cost asphalt 8600 SEK/m Green space length 80 m Asphalt length 60 m

(30)

The price of the DHU comes from an offer from SWEP international AB, they have given a price of 110 000 SEK. Adding this cost to the price of laying down the pipeline, the total investment will be 1 138 000 SEK. From Stockholm Exergi (2021b), the operating and maintenance costs will be 1,3% of the DHU cost. The cashflow of each year will come from sold energy to the grid but as well from the costs that are avoided by self-consuming the heat. The self-consumption saved cost is taken from ME’s pricing in Table 6, this results in a saved cost of 135 784 SEK each year that does not have to be bought from the grid. This will result in a total cashflow of 225 tSEK and 179 tSEK for case 1 and case 2 respectively per year after subtracting the operating and maintenance costs. The net present value will show that after 6-7 years for case 1, and 8-9 for case 2 the investment will be paid off and thereafter turn a profit each year. Hovdestalund will buy heat for a total of 10 375 000 SEK across 25 years. After subtracting heat sold to ME, the final cost will be 9 475 558 SEK for case 2, and

8 564 077 SEK for case 1. Following Kadesjös lifespan of 20 years would result in a total heat bought of 8,3 MSEK. The final costs after sold heat would be 7,4MSEK and 6,49MSEK respectively.

A short sensitivity analysis on how much Hovdestalund can sell their heat to ME for is created for three different prices. Table 10 will show how each case and price point will translate into the change of payback time.

Table 10 Pricing, sensitivity analysis. ME [SEK/MWh] 50 [SEK/MWh] 100 [SEK/MWh] 150 [SEK/MWh] Sold heat [SEK/year] Case 2 44 972 18 607 37 214 55 821 Case 1 90 546 37 105 74 211 111 316 Payback time, years Case 2 8-9 10-11 8-9 7-8 Case 1 6-7 8-9 6-7 5-6

5

RESULTS

Solution 2 was chosen as the technical solution for the district heating case in this study, the solution blueprint can be seen in Figure 10.

(31)

Figure 10 Result: Technical solution 2

In Table 11 the recovery potential is used from Kadesjös report, the facility is using the recovered heat for the following.

Table 11 400kWh heat per cremation across one year.

Total [MWh]

Demand 2020 Sold Bought Self-consumption 852,3 750 643,1 209,2

In Table 12 the recovery potential from different studies that were found during this work is used, the facility is using the recovered heat for the following.

Table 12 242,5kWh heat per cremation across one year.

Total [MWh]

Demand 2020 Sold Bought Self-consumption 852,3 375,8 643,1 209,2

(32)

Figure 11Error! Reference source not found. is showing how much heat is sold and bought while choosing 400 kWh as recovery potential.

Figure 11 Sold/bought heat: 400kWh per cremation.

Figure 12Error! Reference source not found. is showing how much of the recovered heat that is used for self-consumption compared to the heat demand of 2020 while choosing 400 kWh as recovery potential.

Figure 12 Self-consumption: 400kWh per cremation.

(33)

Figure 13 Sold/bought heat: 242,5kWh per cremation.

Figure 14 is showing how much of the recovered heat that is used for self-consumption compared to the heat demand of 2020 while choosing 242,5 kWh as recovery potential.

Figure 14 Self-consumption: 242,5kWh per cremation.

In Figure 15Error! Reference source not found., The heat recovery potential for case 1 and 2 can be seen compared to the heat demand of 2020.

(34)

Figure 15 Demand & heat recovery.

In Figure 16, the heat recovery potential plus bought heat to cover up the demand is shown for case 1 and 2.

Figure 16 Demand, heat recovery & bought heat.

(35)

Table 13. Case 2 will have a payback time of 8-9 years while case 1 will have a payback time of 6-7 years.

(36)

Table 13 NPV

Case 2

Case 1

242,5 kWh

400 kWh

Year SEK Year SEK

0 -1 138 000 0 -1 138 000 1 -968 824 1 -925 830 2 -809 224 2 -725 669 3 -658 659 3 -536 839 4 -516 615 4 -358 697 5 -382 612 5 -190 638 6 -256 194

6

-32 092

7 -136 932

7

117 479

8

-24 421

8 258 585

9

81 722

9 391 703 10 181 857 10 517 286 11 276 324 11 635 761 12 365 444 12 747 529 13 449 519 13 852 971 14 528 835 14 952 445 15 603 662 15 1 046 288 16 674 253 16 1 134 819 17 740 848 17 1 218 339 18 803 674 18 1 297 132 19 862 944 19 1 371 464 20 918 858 20 1 441 589 21 971 608 21 1 507 745 22 1 021 372 22 1 570 156 23 1 068 319 23 1 629 034 24 1 112 609 24 1 684 580 25 1 154 392 25 1 736 981

(37)

Table 13, the values are shown in a diagram instead for easier reading. Figure 17 NPV Case 1 Figure 18 NPV Case 2 -1,500,000 -1,000,000 -500,000 0 500,000 1,000,000 1,500,000 2,000,000 0 5 10 15 20 25 30 SE K Year

NPV Case 1

-1,500,000 -1,000,000 -500,000 0 500,000 1,000,000 1,500,000 0 5 10 15 20 25 30 SE K Year

NPV Case 2

(38)

6

DISCUSSION

The technical solution 2 has several benefits compared to solution 1. Solution 2 has a control unit telling the system that heat is needed for internal use if the circuit gets colder than 50 degrees as used for this example. This creates a partially single circuit which compared to solution one would need two separate circuits. With two separate circuits, it becomes hard to track whether the heat is used internally, sold to the grid, or bought from the grid. This would create further complications when designing the energy meters that will track how much heat is used. With this solution, an energy meter can be placed in both DHU’s. The energy meter in the DHU from the ovens will now only track how much heat is sold to the grid which makes billing a lot easier. The same principal will apply for the second DHU inside the

internal circuit, this energy meter will now only track how much heat is bought from the grid. Looking at Figure 15 & Figure 16 we can see the differences between the daily recovered heat and the daily recovered heat plus bought heat to cover up the remaining demand. Figure 15 has straight lines that represents how much heat that can be recovered compared to the demand. As noticed, the recovered heat does not fully cover the demand for either case when looking at the figures. Daily, the recovered heat of 4,8MWh and 2,91MWh for case 1 and case 2 is not enough to cover up the full day. However, during operating hours, the hourly demand can be met and often exceeded by a lot. Therefore, the recovered heat can be used for self-consumption first and secondly sold. Figure 16 shows another point of view where the recovered heat is added with the bought heat. The facility will only buy heat when the ovens are not running and thus having no heat supply. Together, the recovered heat and buying heat from the grid will fully cover the daily heat demand throughout the year.

The two different cases of choosing whether one cremation will produce 242,5 or 400kWh only has one big difference in how much heat is sold. Since the ovens are always recovering more heat than the facility needs, excess heat will always be sold once they are up and running. Comparing the two cases, Kadesjös idea of 400kWh will turn out to sell roughly double the amount of heat, this seems reasonably since the heat per cremation is almost also doubled. The heat that is bought will always happen during hours and days that the ovens are not running, therefore the bough heat that is needed to supply the facility will remain the same in either case. Same goes for self-consumption, during the hours the ovens are running, the demand will be fully met each time and therefor the self-consumption will be equal for both cases. For both cases, 200 operating days are chosen, this means that a vacation window is assumed during summer between July and August. Looking from Hovdestalund’s

economic perspective, it would be optimal to run the ovens at summer as this would mean more heat used for self-consumption.

Looking at Figure 13 & Figure 11 there is a 4-month gap where there is no sold heat. This is because during this time, ME has an alternative cost that would not make it profitable to buy the heat during these months. The heat will still be sent into the network but without getting paid for it, without a cooling tower this will be necessary as it is the only way to cool down the gases. During the remaining months, the facility can sell heat to ME for the cost of ME’s own

(39)

Self-consumption during these months is still possible and will still benefit Hovdestalund more than selling heat.

Looking at Figure 14 & Figure 12 the self-consumption is seen, the two-month gap is placed due to assumed vacation. Even though Hovdestalund will not get paid during summer months, the produced heat will still need to be supplied to the DH grid. Otherwise, an investment of a cooling tower is necessary anyway to cool down the flue gases.

What is not investigated in this report is how the power output pattern will look like, the mean power output will have an impact on the overall costs. As can be seen in Table 5, a lower power output will also lower the annual costs. To maximize their profit at

Hovdestalund, they should make their power requirements more efficient. This will ultimately also help ME with dimensioning of the power grids.

Looking at investment costs, usually ME will pay for most of the pipeline since this can help with expansion later if needed. For this case, since Hovdestalund will be a prosumer they are assumed to be paying for everything. This will not however be a problem in the long term as can been seen from the NPV result graphs, already by year 6-7 or 8-9 (based on recovery potential) the full investment will be paid off and instead starting to accumulate profit each year. A sensitivity analysis on the discount rate would show that a decrease down to 4% would lower the payback time by roughly 1 year. By instead increasing it up to 8% it would instead increase the payback time of roughly 1 year. The sensitivity analysis on selling price in Table 10 shows a difference of roughly ± 1 year in payback time compared to the payback time of 8-9 and 6-7 years from ME’s set price point. After 25 years the DHU will need a re-investment, but this is a minor cost compared to the accumulated cash flow. The major investment came from the pipeline but that is a one-time cost that does not need replacement anytime soon. Most of the cash flow comes from self-consumption, the facility does not need to buy as much heat from ME. This is the biggest reason to the early payback time for the project.

The result can be compared with the report in subsection 3.4 by Ziemele et al. (2018), the flue gas condenser of 660kW is a similar size to the DHU of 600kW. The flue gas condenser recovers slightly less heat than the crematorium, this is because of the long pipeline of 2000m that reduces the heat recovery with 30MWh monthly. The flue gas condenser recovers roughly 2,25MWh daily compared to this study which has 2,91MWh or 4,8MWh recovered. If the heat loss caused by the long pipeline was removed, the potential rises to 3,1MWh of recovered heat. This might indicate that the case taken from the literature study of 242,5kWh per cremation is more in range compared to Kadesjös 400kWh. Ziemele et al. (2018) receives a payback time of 10,4 years which is close to this study which has 6-9 years in payback time. Ziemele’s study does not account for self-consumption saved cost in their NPV as this study has, this might be a reason for a shorter payback period.

A cooling tower as option is not ideal as it will not benefit Hovdestalund financially. They will not be able to self-consume nor sell excess heat to ME. Therefor they will need to buy heat to cover up the entire heat demand, this results in a yearly cost of roughly 550 407 SEK. As can be seen from Table 1, by not having a cooling tower that is cooled by water instead of air, this will also reduce the water usage by a lot.

(40)

Comparing the results with Kadesjös report, a slight difference in investment cost is noticed. The difference could come from either the pipeline or DHU cost. Looking at the final costs for both reports, this report shows a lower overall cost across 20 year. Kadesjös receives a total cost of 10MSEK while this report shows a total of 7,4MSEK and 6,49MSEK based on recovery potential. Kadesjös concluded that geothermal would save 1M compared to the heat recovery method, geothermal would cost 8,8M after 20 years. This would mean that from the results of this report, the heat recovery method would be the optimal solution as it shows a lower cost of 1-2MSEK after 20 years.

A complication when it comes to who owns what in the system, this could affect the costs for Hovdestalund. If ME would own the entire recovery circuit, this would take responsibility off from maintenance and similar. From ME’s perspective, having to maintain every part of the DH grid would take a lot of time. There is also a risk that with a different demarcation it could affect the customer's operations to an extent that is not desirable. The system will be divided from the DHU’s, the system behind the DHU (own circuit and oven-circuit) will be owned by Hovdestalund while the primary side will be owned by ME.

The crematorium will supply the DH network with roughly 92 degrees water, this is

considered high grade heat and thus is also good for the network. When looking at water that is around 70-75 degrees, this is around the point to where the water is only diluted into the network. Connecting many facilities that deliver this low temperature water to the network might eventually cause problems if the amount becomes big enough. With few facilities, this could be considered as “charity” from ME as the facilities does not need to invest in

additional cooling towers.

In the two different DHU’s there will be two different HX, this is because the ovens are recovering higher heat than the internal circuit will ever need to buy. Inside the oven-DHU, the HX will be designed to manage a load of 0,6MW hourly if two ovens are running

simultaneously. The internal-DHU will only need to be designed for 0,3MW hourly since this is a bit higher than the maximum heat demand across the year of 2020. The internal DHU already exists and does therefore not need an investment.

In Figure 9, the temperatures for each side of the HX are shown. The HX is assumed to lose 3 degrees while transferring the heat between the circuits. While a better HX could be designed to lower the temperature to around 1-2 degrees, this would also result in a way higher

investment cost that could make the whole investment not worthwhile. This temperature can be further investigated to both being increased and decreased to find the optimal

temperature when looking from an economic perspective. For this study, 3 degrees will be used throughout the entire project. The HX also needs to consider the inlet temperature to the ovens, by supplying to cold water, the ovens might not reach the desired temperature due to it being cooled down by the supply stream. By checking with Patrik Selinder at FVB, there would be no problem with 50 degrees, the water could be in a range of 40-60 degrees with little to no effect on the ovens. When designing the DHU, they are designed for summer-condition, at summer the temperature is higher and thus creating lower heat transfer for the HX.

(41)

CONCLUSIONS

To simplify the cooperation between Mälarenergi and Hovdestalund, a technical solution which can separate when heat is supplied to the network versus withdrawn will help with billing and general surveillance across the system. The production/consumption pattern will stay the same for both recovery cases 1, and 2. The only difference between the two is the amount of heat that will be sold to the grid. Hovdestalund will be able to fully cover their heat demand on the hours that the ovens are up and running, by self-consuming the heat during these hours will reduce the annual cost by 136 tSEK. The excess heat can then be sold to Mälarenergi’s district heating grid to be used elsewhere where it is needed. Hovdestalund can sell heat for somewhere between 45-90,5 tSEK annually based on how much heat that can be recovered from the system. The recovery system will not affect Mälarenergi’s grid or

production to the point of having a negative impact. The recovered heat will count as high grade at roughly 92 degrees, being lower at around 70-75 degrees could cause some problems with temperatures. Even then at 70-75 degrees there would need to be an extreme amount of production that will not be possible from only Hovdestalund alone. During June-September, Mälarenergi has negative production prices. Therefore, in these months, Hovdestalund will not be able to sell heat to the grid for profit.

Investing in a heat recovery system is a worthy investment, by using the flue gases to heat up water to produce high grade heat for the district heating will become more viable in the future for any process. Hovdestalund can create future profit by incorporating heat recovery to the ovens. The daily recovered heat of 4,8MWh and 2,91MWh for case 1 and case 2

respectively will show that the payback time will occur after 6-7 or 8-9 years. Comparing this to Kadesjös result on geothermal, this recovery system can save up to two million SEK after 20 years. Using a cooling tower as a solution is not profitable either as it will not be able to sell nor self-consume the heat.

7

SUGGESTIONS FOR FURTHER WORK

For future work, a deeper investigation on how much Hovdestalund should sell the recovered heat to Mälarenergi for. As of now, the price is assumed to be equal to the CHP plant fuel costs. Managing power output is something that is becoming more relevant, recently

Mälarenergi changed their payment method based on this. Therefore, finding a more power efficient way could affect the crematoriums annual costs positive, as well as helping

Mälarenergi with dimensioning of the power grid. A sensitivity analysis on the heat exchanger could also help with optimizing the potential and a sensitivity analysis on the discount rate could also be investigated.

(42)

8

REFERENCES

Apana. (2021). What Is a Cooling Tower and How Does It Work? Retrieved from Apana:

https://www.apana.com/what-is-a-cooling-tower-and-how-does-it-work/#:~:text=Air%20circulates%20through%20the%20cooling,equipment%20or% 20air%20conditioning%20system.

Cronholm, L.-å., Grönkvist, S., & Saxe, M. (2009). spillvärme från industrier. Svensk fjärrvärme. Retrieved from

https://energiforskmedia.blob.core.windows.net/media/1202/spillvaerme-fraan-industrier-och-lokaler-fjaerrsynsrapport-2009-12.pdf

Dalenbäck, J.-O., Lennermo, G., Andersson-Jessen, P.-E., & Kovacs, P. (2013). Solvärme i

fjärrvärmesystem. DiVA.

doi:http://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1058766&dswid=1819

Foley, A. (2013, September 11). How to Model a Shell and Tube Heat Exchanger. Retrieved from Comsol: https://www.comsol.com/blogs/how-model-shell-and-tube-heat-exchanger/

Gunnar Lennermo; Patrick Lauenburg. (2016). Feed-in distributed solar thermal plants in

district heating systems. DiVA. Retrieved from

https://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1055474&dswid=5121

Harnesk, T. (2015, Januari 20). Vattenfall använder fjärrvärme från krematorier. Retrieved from NyTeknik: https://www.nyteknik.se/energi/vattenfall-anvander-fjarrvarme-fran-krematorier-6395877

Hellborg, J. (2017). Modelling of shell and tube heat exchangers. Lund University. Retrieved from

https://lup.lub.lu.se/luur/download?fileOId=8900249&func=downloadFile&record OId=8900243

Johansson, P. (2021, 02 24). (C. Lindquist, & J. Ericson, Interviewers) Kadesjös. (2020). Energiutredning Hovdestalund. Västerås.

Lennermo, G., & Lauenburg, P. (2016). Feed-in distributed solar thermal plants in district

heating systems. DiVA. Retrieved from

https://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1055474&dswid=5121

Lennermo, G., Lauenburg, P., & Brange, L. (2016). Små värmekällor. Energiforsk. Retrieved from

https://energiforskmedia.blob.core.windows.net/media/21287/sma-varmekallor-energiforskrapport-2016-289.pdf

Levy, E., Bilirgen, H., & DuPont, J. (2011). RECOVERY OF WATER FROM BOILER FLUE

Figure

Figure 1 DHU with primary and secondary circuit
Figure 2 Differential pressure curve
Table 1 Water requirements for different types of plants (World nuclear association, 2020)
Table 2 Kadesjös input variables.
+7

References

Related documents

for Control of the Flue Gas Desulfurization System in Power Plants. Experimental and Quantum

För att förstå och kunna bemöta dem på ett värdigt sätt är det viktigt att ha kunskap om vilka följder långvarig smärta kan ha på människors aktivitet i det dagliga

Människor med psykisk ohälsa har ofta en obalans i sitt aktivitetsmönster (31) och kan gynnas av meningsfulla aktiviteter dagtid, exempelvis i trädgårdsmiljö, för att kunna få

en effektiv teknisk åtgärd. I detta direktiv avses med teknisk åtgärd varje teknik, anordning eller komponent som har utformats till att vid normalt bruk förhindra eller begränsa

The channel dependence could not be measured using the new method due to the frequency dependence. The spectrum analyzer cannot handle more than one frequency setting simultaneously.

In other words if the output increases above the reference level the controller will increasing the speed of the lime motor to suppress the output error, but at the same time

The sea velocity is 3 meters per second for the whole year, the distance analyzed before making the local energy balance is 50meters, the conductivity of the steel of the

If the temperature of the gases is 50ºC instead of 55ºC, and the moisture content is high, almost 10kW more can be recovered with almost the same area. Furthermore, if the