Applied Energy 279 (2020) 115883

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Connecting circular economy and energy industry: A techno-economic

study for the Åland Islands
Kirsikka Kiviranta *, Tomi Thomasson , Jonne Hirvonen , Matti Tähtinen
VTT Technical Research Centre of Finland Ltd., Koivurannantie 1, 40100 Jyväskylä, Finland

H I G H L I G H T S

• Coupling renewable energy to circular economy in a case region has been analysed.
• Circular economy has potential to manage high variable renewable energy outputs.
• Processing waste materials can increase the value of renewable energy investments.
• Establishing a circular ecosystem requires extensive cooperation between companies.

A R T I C L E I N F O A B S T R A C T

Keywords: Energy plays an essential role in circular economy because circular activities such as material processing require
Circular economy power and heat. In parallel, the rate of the transition to renewable energy is not adequate to meet the increasing
System optimization energy demands. The objective of the study is to evaluate whether circular economy could increase the value of
Variable renewable energy
variable renewable energy investments and hence accelerate the transition towards renewable energy. The study
involves a combined energy system and material flow analysis. The study is performed on a selected case region
as the processes in circular economy and the availability of renewable energies are always local and depend on
regional conditions. The Åland Islands was used as a case platform in the study as the electricity generation
capacity from wind power in the region is expected to increase significantly in the near future resulting in high
variability in the local power supply. Four alternative scenarios are analyzed in which the variable regional
renewable energy supply exceeding the local demand is integrated for different purposes: power exports, circular
economy, partly electrified transportation sector and district heating. With the highest annual system net profit
(0.72 M€), integrating the power production peaks of variable renewable energy into circular economy was
found to outweigh the annual economic benefits of power exports (− 0.43 M€), the partly electrified trans­
portation sector (− 0.50 M€) and district heating (− 0.27 M€). Therefore, the value obtained from the products
derived from circular processes increased the value of the renewable energy system and would hence promote
investments in renewable energy in the region.

1. Introduction renewable energy production can be maximized is vital to ensure suf­

ficient progress in the energy transition [7]. However, as the output of
The global targets of the Paris Agreement [1] and national policies variable renewable energy sources does not match the power demand,
concerning climate change mitigation together with significant cost the regional large-scale deployment of renewables, primarily intermit­
reductions in variable renewable energy (VRE) [2,3] are driving a tent wind and solar power, will increase the need for energy system
transition from fossil-based energy systems to ones dominated by flexibility [8,9]. Flexibility in the energy system is necessary to avoid the
renewable energy throughout the world. Although the net capacity curtailment of excess VRE supplies and to increase the adaptability of
addition of renewables has increased every year for nearly two decades other power generators during periods of low wind and solar output.
[4], the current deployment rate of renewables is not fast enough to Alternative measures to increase energy system flexibility to manage
meet the set long-term climate goals [5,6]. Directing renewable energy VRE integration have been reviewed. At the core of the alternative
investments to regions where capacity factors and hence energy yields of measures to increase energy system flexibility to manage VRE

* Corresponding author.

https://doi.org/10.1016/j.apenergy.2020.115883
Received 16 January 2020; Received in revised form 1 September 2020; Accepted 12 September 2020
Available online 30 September 2020
0306-2619/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
K. Kiviranta et al. Applied Energy 279 (2020) 115883

integration are energy storage, demand side management and different feasibility of integrating excess VRE production into circular economy
power-to-X strategies, in which excess electricity is converted to thermal processes in contrast to three more conventional energy system flexi­
energy, gas, hydrogen or used as a charging source in the electrified bility alternatives including power exports to neighboring power areas,
transportation sector [10,11]. An alternative approach to utilize excess power-to-electric vehicle charging in a partly electrified transportation
electricity is introduced in [12], where low-cost electricity generated by sector and power-to-heat conversion via electric heaters for district heat
intermittent energy sources is suggested for producing energy-intensive production. The Åland Islands are selected as a case platform for the
intermediate products such as pure water, chemicals, alloys and com­ study due to its substantial upcoming VRE investments and the antici­
posites to create additional value for renewable energy investments. By pated resulting variance in the regional power supply [22] and also
processing regional side and waste material streams to become inter­ because the island region is a compact geographical area enabling clear
mediate high-value products, circular economy could provide an addi­ system boundaries for both the energy system and circular economy.
tional power demand at times when the regional power supply exceeds The study is performed using a developed mixed integer linear
local demand. The potential conjunction between circular economy and programming-based model, in which the energy system and material
the resulting low-cost peak power supply from intermittent power flow optimization are combined. The circular ecosystem on the Åland
sources could thus help accelerate the transition to renewable energy. If Islands is illustrated by identifying local industries which generate side
a company is able to attain additional value by processing their side and and waste material streams that could financially benefit from value-
waste material streams into intermediate high-value products during adding material circularity.
periods of low-cost power supply, the increased material value created
by VRE could create incentives for companies themselves to invest more 2. Methodology
in renewable energy.
Islands provide an interesting subject for VRE integration studies. The methodology of this study is divided into four main sections.
Most islands are highly dependent on imported energy although islands First, a short description of the modelling framework applied in this
are often abundant in VRE resources such as wind and solar irradiation study is presented. Second, a reference scenario of the energy system of
[13]. In addition to initiatives to increase energy self-sufficiency, the the Åland Islands for the year 2025 is described. Third, three alternative
energy system structures on islands are often simpler in relation to energy system scenarios are constructed for the year 2025 including
continental regions [14]. With clear system boundaries and their small alternative utilization pathways for the power supply exceeding local
scale, islands provide a feasible platform to assess and implement new demand. Fourth, cost and market assumptions for the scenarios are
measures to enable the wider adoption of VRE and can therefore also act presented.
as blueprints for energy transition in larger continental areas [15].
Flexibility measures to enable higher VRE penetration in island regions 2.1. The optimization model
have been evaluated in several case studies. Techno-economic analyses
for alternative renewable energy system configurations have been con­ The BRAIN framework, created using Pyomo [27] and CPLEX [28],
ducted for La Gomera (Canary Islands, Spain) [16], where the flexibility and initially presented in [22], enables the creation of a hourly-resolved
potential was recognized in bidirectional vehicle-to-grid charging, in combined capacity planning and dispatch optimization model based on
which electric vehicles acted as mobile energy storage devices on the mixed integer linear programming. The model determines the optimal
island. The importance of transmission and flexible demand in VRE dispatch schedule of power and heat production units, storage systems
integration in the Canary Islands archipelago has been highlighted along and power transmission and resolves the optimal capacity of the po­
with coupling energy, transportation and heating sectors with electric tential dispatchable unit investments while respecting defined con­
vehicles and electric heaters [17]. A pumped hydro storage plant was straints. The objective function of the model aims to maximize the
discovered to be a techno-economical solution to increase wind and system operating profit, which includes income from sold electricity,
solar energy penetration in the Faroe Islands (Denmark) [18]. As for district heat and exported electricity, as well as costs from used fuels,
practical examples of renewable energy islands with high VRE pene­ OPEX, CO2 emission allowances, imported electricity and annual in­
tration, a hybrid system consisting of a wind farm and a pumped hydro vestment payments from optimal new investments. In this study, the
storage has been implemented on El Hierro (Canary Islands, Spain) [19], model was expanded to optimize energy systems in conjunction with
whereas a wind turbine-photovoltaic-electrochemical battery hybrid has selected circular processes. Unit models of circular processes comprising
been realized in the islands of Tilos (Greece) [20]. essential process steps were included in the optimization model, and the
The Åland Islands are an autonomous region of Finland located in objective function was set to include operating costs and incomes from
the Gulf of Bothnia, and have been used as a case platform in several the selected circular processes. A more thorough description of the
renewable energy system studies on flexibility as the islands aim to model formulation is presented in Appendix A.
become entirely renewable in regards to energy by 2025 [21]. The
current energy system of the Åland Islands including both the power and 2.2. Establishment of a reference scenario for 2025
thermal heat sector is described in [22]. To achieve energy autonomy,
the Åland Islands intend to increase their installed wind power capacity The reference scenario in this study represents the energy system of
almost nine-fold from the current 21 MW [23] to approximately 185 the Åland Islands in 2025 in which the demand for power [29] and
MW [24] in the upcoming decade. The significant increase in VRE ca­ district heat [30] have grown according to the consumption growth
pacity will result in high fluctuations in the local power supply. Pro­ trends of the past decade. The scenario also includes the planned in­
posed solutions to promote high-level VRE integration in the Åland vestments in additional renewable capacity. An hourly time series of the
Islands include, for instance, a highly electrified transportation sector power and district heat demand for the year 2017 was provided by the
[25] along with heat storage and synthetic fuel production using power- local transmission system operator Kraftnät Åland and by the local en­
to-gas technologies [14] and flexible combined heat and power gener­ ergy company Mariehamns Energi. Both data series were scaled result­
ation from domestic biomass [26]. ing in annual power and district heat demands of 350 GWhe and 132
Based on the literature review, circular economy has not been GWhth for the year 2025, respectively. Thermal losses of district heat
considered as a demand source to utilize the excess power supply from (10.9%) were kept constant with the values from 2017 [31]. The
renewable energy on the islands. In this study, excess power supply transportation fuel demand for gasoline and diesel for 2025 were also
refers to power that is produced by VRE sources in the case region, but scaled based on the consumption growth trends of the past decade
which cannot be utilized locally during the given time frame. The Åland resulting in fuel demands of 111 GWh and 112 GWh for gasoline and
Islands are utilized as a case platform in the study to evaluate the diesel, respectively [32].

2
K. Kiviranta et al. Applied Energy 279 (2020) 115883

The wind power capacity in the reference scenario was increased represented approximately 85% of the fuel used for district heat pro­
from the current 21–185 MW based on the regional wind power projects duction in the main district heating network of the Åland Islands [30].
in the pipeline [24]. The fleet was divided into two categories of onshore Smaller heating networks in the region are also fueled by residues from
38 MW (30 MW) and offshore 147 MW (134 MW) turbines based on the the local forest industry [41] and partly by biogas from a local milk
location of the existing [23] and planned [24] turbines. The values in the processing plant [42]. The local chip factory [43], milk processing plant
brackets represent new investments in the mentioned turbine type. Wind [42] and wastewater treatment plant [44] all own individual biogas
speed data [33] from the closest observation point for each turbine was production units, in which biodegradable side products are converted
converted to the expected hub height by applying wind power law. The into energy applied mainly in the internal processes of the companies
wind speeds were converted into wind power output by applying wind themselves. A land-based fish farm plans to build a biogas production
power curves from a common supplier of existing onshore [34], existing unit to convert wastewater sludge from fish cultivation water and resi­
offshore [35] and new wind turbine investments [36]. The solar PV dues from fish slaughtering processes to biogas for internal energy use
capacity on the island was only marginal in 2017 [37] and hence the [45]. The recirculating aquaculture system used in the fish farm, in
output is not observed on a system level. As for the 2025 scenario, a total which cultivation water is continuously purified and reused, could also
solar PV capacity of 15 MW was assumed to exist comprising of resi­ provide a potential demand source for oxygen to optimize fish growth.
dential solar PV installations. The solar PV output was calculated based Furthermore, a mobile wastewater treatment unit is being piloted in the
on solar irradiation data from 2017 [33] and a constant panel efficiency Åland Islands to extract suspended solids, phosphorous, nitrogen and
of 14%. carbon from wastewater of a fish processing plant [46]. In addition, fish
In the 2025 scenario the district heat generation capacity includes residues from local fish farming have been applied in small-scale bio­
the existing 11 MW and additional 5 MW heat production units fueled diesel production largely utilized by a local bus company [47].
with solid biomass [30] as well as heavy fuel oil fueled transportable In 2016, the total amount of generated waste in Åland Islands was
production capacity, serving as a back-up reserve [31] totaling 45 MW approximately 45,600 tons, consisting of 42,600 and 3000 tons of non-
[30]. The current district heat generation capacity in the Åland Islands hazardous and hazardous waste, respectively [48]. The content of the
also includes two CHP units fueled by heavy fuel oil [30]; however, these non-hazardous waste is depicted in Fig. 1. The majority of the waste
units are expected by be decommissioned by 2025 and were therefore treated in the Åland Islands was biodegradable waste, which was largely
not included in the simulations. A constant thermal unit efficiency of composted. Household and similar waste was transported to Finland and
90% was assumed for both biomass boilers whereas the efficiencies of Sweden for thermal waste incineration. Recyclable waste including
the oil boilers applied as a back-up reserve are expected to be 86%. paper, metal, glass, plastic, and rubber was mainly transported to be
The hourly electricity price data in the Nord Pool SE3 market area in treated outside the islands. In addition, all hazardous waste was
2017 [38] was applied to represent the market conditions in 2025 in exported. As the waste volumes of the exported recyclable material
Åland Islands. The power transmission was aggregated into a single 180 flows are relatively small, they are not considered as potential waste
MW interconnector [39] to reduce the model complexity. In reality, the streams for upgrading processes in this study. Furthermore, the avail­
Åland Islands is connected to the market areas of Finland and Sweden able data related to composition of the recyclable waste streams, such as
with separate power transmission lines [40]. plastics, was considered insufficient for an in-depth analysis of the po­
tential utilization pathways. Therefore, in this study, the biodegradable
2.3. Establishment of the studied scenarios waste streams are at the core of the circular ecosystem illustrated for the
Åland Islands.
Three alternative scenarios along with the reference scenario are Fish residues, are an untapped resource in Åland Islands as they are
analyzed in this study to evaluate the potential of circular economy to currently mainly treated with formic acid and sold as fur animal feed to
increase the value of a renewable energy system. The studied scenarios northern Finland [50]. The annual amount of fish residues produced in
were: (1) the power-to-circular economy (P2CE) scenario; (2) the power- the Åland Islands can reach up to 3000 tons including fish residues from
to-electric vehicles (P2EV) scenario; and (3) the power-to-heat (P2H) local fish farming and from local fish processing operations where both
scenario. The three alternative scenarios were built on top of the refer­ locally farmed and imported fish are processed [47]. The majority (99%)
ence scenario, and hence all parameters related to the energy system of fish farmed in the Åland Islands are rainbow trout and their residues
itself remained the same as in the reference scenario unless otherwise consist of fat (60%), protein (20%) and nutritious water (20%) [47].
stated. In the four scenarios (including the reference scenario), the VRE According to the composition of the fish waste stream, this study as­
supply exceeding local demand may be used for different purposes. In sumes the annual yield of waste oil and waste protein fractions from fish
the reference (Base) scenario, power peaks are exclusively exported to residue to equal 1800 and 600 tons, respectively. In this study, the
the neighboring power areas of Sweden and Finland. The excess power separated oil fraction was applied in biodiesel production, whereas the
supply may also be exported in the three alternative scenarios. However, extracted protein fraction was used as a substrate for biogas production.
in the P2CE scenario, the excess power supply may also be applied in the The upgrading process of fish waste begins with a fractionation step, in
circular processes defined later. In the P2EV scenario, the excess supply which formic acid, electricity and heat are applied to separate waste oil
can be used for charging electric vehicles whereas in the P2H scenario, and other fish remains including nutritious water and protein fractions
the excess power supply can be converted into district heat via electric from the residue [51]. The waste oil is further processed into biodiesel
boiler. The detailed construction of each alternative scenario is pre­ with methanol, an alkaline catalyst, electricity and heat resulting in
sented in the following paragraphs. output streams of biodiesel and crude glycerin [51]. The upgrading
To define the circular processes for excess power supply utilization in process takes place in a single unit, in which fish residue from individual
P2CE scenario, the material flow characteristics of the Åland Islands had farmers and operators on the island are collected. In other scenarios than
to be evaluated to detect potential waste streams applicable for value- P2CE, including the reference scenario, the residue was sold as animal
adding material processing. The evaluation was conducted by review­ fur feed with an associated price of 150 €/t [47], instead of processing it
ing official statistics, the research literature and publicly available data into biodiesel.
from companies located in the Åland Islands and hence the illustrated An investment in a centralized biogas plant has been discussed in the
circular ecosystem could have potential for practical implementation. Åland Islands [52] and hence all scenarios, including the reference
Several circular processes already exist in the Åland Islands. Side scenario, assume that an investment in a biogas unit converting avail­
products such as branches, tops, bark and sawdust from the local forest able biodegradable waste in the Åland Island into biogas has already
industry are already exploited in domestic district heat production [41] been made. The biogas potential has been studied in the Åland Islands
and as of 2017, residues from local wood processing operations [53] and the methane yield of the referred study (2.06 million m3) is

3
K. Kiviranta et al. Applied Energy 279 (2020) 115883

Fig. 1. Breakdown of non-hazardous waste material streams in the Åland Islands in 2016. Data derived from [49].

applied to evaluate biogas supply in the scenarios excluding the P2CE peak production periods of VRE in the Åland Islands is expected to in­
scenario. For the purpose of establishing circular processes in the P2CE crease the profitability of these processes. Therefore, an alkaline elec­
scenario, some modifications were made to the biodegradable waste trolyzer (2.6 MWe) with 67% efficiency [60] was combined with the
supply of the given study based on findings from the literature. The biogas plant. The capacity of the electrolyzer was determined based on
annual amount of fish waste in the biodegradable waste feed-in was the maximum amount of electricity required to produce hydrogen
increased from 560 t/a to 600 t/a in accordance with [47], increasing capable to stoichiometrically convert all the available carbon dioxide
the annual methane yield of fish waste by 15,000 m3 by applying dry (0.288 t/h) from raw biogas into biomethane in a Sabatier reaction
substance and methane yield estimates from [53]. Moreover, five mobile [57,58]. Before the methanation process, carbon dioxide and methane
wastewater treatment units are expected to be piloted in the Åland from the raw biogas feed are separated, resulting in purified fractions in
Islands to partially purify wastewater from fish slaughtering and pro­ both streams. The purified fraction of carbon dioxide, upgraded to
cessing operations. The dry matter fraction from the wastewater treat­ biomethane with hydrogen, is later referred to as synthetic natural gas
ment units is expected to increase the annual methane yield by 56,300 (SNG). The combined feed of the purified methane and the generated
m3. An additional biodegradable waste stream was also assumed to be SNG is later referred to as biomethane which is applicable for use as a
derived from a local brewery, which tripled its annual production ca­ transportation fuel in parallel to methane. The generated oxygen in the
pacity in 2017 to 2.5 million liters of beer [54]. Currently, the mash from electrolysis could be partially applied in the land-based fish farm located
the brewery is fed as animal feed to local cattle and pigs. In this study, in the Åland Islands and the derived process heat from the electrolyzer
the mash is instead utilized as a biodegradable waste fraction for biogas could be applied as district heat. An intermediate, short-term com­
production. The estimated annual mash yield of 438 tons with a water pressed hydrogen storage was added to the system to partly decouple the
content of 75% [55] is expected to result in an additional methane yield hydrogen production from the biological methanation process.
of 43,800 m3. With given modifications, the annual methane (CH4) yield Eventually, the circular processes selected to utilize excess power
was increased from 2.06 million m3 [53] to 2.18 million m3. As the supply in Åland Islands were fish waste fractionation and biodiesel
methane concentration in the raw biogas is expected to remain constant production along with raw biogas upgrading through a biological
(63%) to the referred study [53] the hourly raw biogas supply of the methanation process. Therefore, simplified unit models of fish waste
biogas plant is estimated to be 395 m3 throughout the year. fractionation, biodiesel production, electrolysis and biological metha­
Besides some impurities, the remaining fraction of biogas (37%) nation were included in the optimization model as additional power
consists of carbon dioxide (CO2) [56]. The CO2 fraction obtained from demand nodes. With respect to the circular processes, the model is able
the generated biogas could be upgraded to biomethane in a methanation to optimize alternative options on an hourly basis. In the P2CE scenario,
process with hydrogen derived from an electrolyzer. Hydrogen pro­ the model optimizes whether biogas is upgraded to biomethane or sold
duction by water electrolysis and the subsequent conversion of as a raw biogas, and if fish waste is fractionated and converted to bio­
hydrogen and carbon dioxide to methane is considered a promising diesel or sold as animal fur feed instead of the conversion to a higher-
option to utilize low-cost power supply during peak production periods value product. A visual representation of the studied circular processes
of fluctuating VRE generation [57,58]. As the electricity price is in the P2CE scenario is shown in Fig. 2 along with the most essential
considered an important factor in the economic feasibility of electrolysis process parameters. Available fish waste volumes and the subsequent
and methanation processes [59,60], the low-cost power supply during fractions of oil and protein are derived from [47] whereas the

4
K. Kiviranta et al. Applied Energy 279 (2020) 115883

Fig. 2. Simplified material flow chart for the illustrated circular ecosystem for the Åland Islands. Output streams of the circular ecosystem are outlined with a dashed
line. The orange line represents the additional power demand nodes added to the system. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)

parameters related to fish waste fractionation and biodiesel production

are derived from [51]. The output streams of the circular ecosystem are Table 1
Main supply and demand technologies applied in the scenarios. Values in
outlined with a dashed line.
brackets represent new wind power investments.
Whereas the P2CE scenario is based on actual material flow volumes
in the Åland Islands, the following P2EV and P2H scenarios are based on Technology Unit Scenario

simpler assumptions with the aim of providing a basis to compare the Base P2CE P2EV P2H
feasibility of circular economy in managing high VRE outputs. In the Wind onshore MWe 38 (30) 38 (30) 38 (30) 38 (30)
P2EV scenario, electric vehicles are considered as an additional power Wind offshore MWe 147 147 147 147
demand node in the system. Assuming that 25% of the car stock of the (134) (134) (134) (134)
Åland Islands in 2017 [61] would be replaced by electric cars, 5960 Solar PV MWe 15 15 15 15
Biomass boiler #1 MWth 11 11 11 11
electric vehicles will be in use by 2025. The electric vehicles were Biomass boiler #2 MWth 5 5 5 5
assumed to consume 0.2 kWh per kilometer [62]. With daily road Oil boiler MWth 45 45 45 45
transportation demand for each car at 41 km [63], the estimated daily Transmission MWe 180 180 180 180
power demand for one electric vehicle equals 8.2 kWh. In the model, capacity
Fish waste kWe/ 8.4/
charging was set to be allowed in a six-hour period between the hours of
fractionation kWth 48.2
23 and 05 with a maximum hourly charging rate of 3.8 kW [64]. Biodiesel production kWe/ 9.4/7.2
Charging was set to be possible from the grid to the vehicle without a kWth
possibility for bidirectional vehicle-to-grid charging. In the P2H sce­ Electrolysis MWe 2.6
nario, an electric boiler was added to the modelled system to provide an Methane yield million 2.06 2.18 2.06 2.06
m3
additional demand source for the excess VRE supply and an alternate Electric vehicles MWe 23
option for heat generation in the district heating network. For the boiler, Electric boiler MWe 10
a capacity of 10 MW and thermal efficiency of 95% was selected. The
scenario-wise main supply and demand parameters are shown in
Table 1. comparison to biomethane, the price of biogas was assumed to be lower
(60 €/MWh). As the energy content of biodiesel (11.5 MWh/t [51])
equals the energy content of diesel (11.5 MWh/t) the income from
2.4. Cost and market assumptions applied in the scenarios biodiesel was expected to be equivalent to consumer diesel prices
excluding value-added tax (2017: 0.98 €/l [68]).
Investment costs for new power production technologies applied in
all scenarios are presented in Table 2. The table also includes investment 3. Results and discussion
costs related to the circular processes covered in this study. For all the
investments, a weighted cost of capital of 7% is applied to determine the In this section, the cost-optimal operation and the value of the
annual investment payment. The electrolyzer applied in the P2CE sce­ studied scenarios are evaluated. Annual power and district heating
nario is equipped with a hydrogen compressor with an investment cost structures for the four scenarios are analyzed in addition to the annual
of 230 €/kWe [65] of electrolysis input power. The hydrogen tank with cost breakdown of the studied scenarios. A detailed cost analysis is
an investment cost of 13.5 €/kWhH2 [65] is scaled to store one hour conducted for the value-adding products in the circular ecosystem of the
production of hydrogen at full load. P2CE scenario. Furthermore, a sensitivity analysis is carried out to
The main market parameters applied in the scenarios are presented evaluate the impact of alternate product prices (biogas versus bio­
in Table 3. The price of biomethane was set at an optimistic value of 80 methane and fish waste versus biodiesel) on the profitability of the
€/MWh [60]. As the methane content in raw biogas is lower in

5
K. Kiviranta et al. Applied Energy 279 (2020) 115883

Table 2
Investment parameters applied in the scenarios for power generation and circular economy technologies.
Process CAPEX Fixed OPEX [% of CAPEX] Lifetime (a) Source

Onshore wind 975 €/kW 2.3 30 [65]

Offshore wind 1300 €/kW 2.3 30 Assumption based on [14]
Solar PV 715 €/kW 1.4 30 [65]
Electrolysis 12,301 €/kW 2.5 20 [60,65]
Biological methanation 730 €/kWSNG 2.5 20 [60]
Fish waste fractionation 420 €/tfish waste 2.5 15 Assumption based on [47]
Biodiesel production 300 €/tbiodiesel 2.5 15 Assumption based on [66]
Hydrogen storage 13.5 €/kWh 0.14 30 [65]
Electric boiler 65 €/kW 1.6 20 [67]
1
Includes compressors for hydrogen storage.

vehicles within the given daily charging period. Similarly, in the latter
Table 3
scenario, the biodiesel production from fish waste was optimally oper­
Main market parameters applied in the scenarios.
ated continuously, increasing the demand for imported electricity at
Parameter Cost Source times when the local supply was not adequate to meet both the local
Electricity (average) price 31.2 €/MWh [38] power demand and the power demand of the additional circular process.
District heat price 80 €/MWh [69] As seen in Table 4, the added circular processes in the P2CE scenario
Biomass price 22 €/MWh [70]
increased the local power consumption in the Åland Islands by 3.9% in
Oil price 45 €/MWh [14]
CO2 emission allowance price 30 €/t [71]
comparison to the Base scenario. The increase in power consumption
Biogas price 60 €/MWh Assumption based on [60] depended on the energy intensity of the selected circular processes.
Biomethane price 80 €/MWh [60] Consequently, in the P2H and P2EV scenarios, more local production
Biodiesel price 107 €/MWh [68] was integrated in the energy system, as the share of power exports was
Fish waste price 150 €/t [47]
lower than in the P2CE scenario. Therefore, the power demand of
Formic acid price 600 €/t [60]
Methanol price 390 €/t [60] electric vehicles or heat conversion from the excess VRE supply would
Glycerin price 200 €/t [51] provide a better option to increase the integration of the local power
supply in comparison to the circular ecosystem with the defined system
constraints.
circular ecosystem in which biomethane and biodiesel are produced The fuel consumption for district heat production reduced in the
from the applied waste streams. P2CE and P2H scenarios compared to the Base scenario. In the P2CE
The system optimization results show variation between the studied scenario, the electrolyzer contributed to district heat production by
scenarios in annual power consumption, import and export volumes, as supplying process heat to the district heating network, whereas in the
well as in annual district heat generation quantities with the associated P2H scenario, the electric boiler converted the excess power supply to
fuels as seen in Table 4. In accordance with the simulation results, the district heat. In the P2H scenario, the oil consumption decreased
local power generation in the Åland Islands is dominated by variable significantly (13.6 GWh and 56%) while the biomass consumption
wind power (611 GWh/a) whereas the role of solar PV (14 GWh/a) is remained constant being the less expensive district heat production
only supplementary. Consequently, the amount of generated electricity method. The reduction in oil consumption was high in the P2H scenario,
was almost twice the power consumption (350–370 GWh) of the Åland as the electric boiler was able to produce heat independently of other
Islands in all scenarios. In comparison to the Base scenario, the annual processes, allowing the reductions to focus on the peak production of
power consumption increased in the P2CE, P2EV and P2H scenarios as district heat with oil. In the P2CE scenario, the process heat derived from
additional power demand nodes were added to the analyzed systems. the electrolyzer reduced both oil consumption (1.9 GWh or 2%) and
The integration of the local power supply to the Åland Islands increased biomass consumption (2.0 GWh or 1.8%). The achieved reductions were
in the P2CE, P2EV and P2H scenarios as the amount of power exported smaller than in the P2H scenario as the process heat availability of the
decreased in comparison to the Base scenario. The amount of imported electrolyzer was dependent on the downstream processes: if biogas is not
electricity increased in the P2EV scenario and slightly in the P2CE sce­ upgraded, hydrogen production is not required. It should be also noted
nario. In the former scenario, the availability of wind and solar power that the maximum hourly potential of heat derived from alternative
was not sufficient at all times to meet the power demands of electric sources rather than biomass and oil were notably smaller in the P2CE
(0.9 MW) scenario in relation to the P2H (9.5 MW) scenario, which led
Table 4 to a difference between the scenarios.
Power and district heat generation structures of the simulated scenarios. Change The annual income and cost breakdown of the scenarios is presented
(%) describes the relative change of the value compared to the Base scenario. in Fig. 3, with the structure being the same in all the scenarios. The
Parameter Base P2CE P2EV P2H annual income was formed from locally sold and exported electricity,
Power consumption (GWh) 350 364 370 364
sold district heat, and sold fuels. Annualized investment costs were the
Change (%) 3.9% 5.0% 4.1% largest contributor to the annual system costs. In addition, costs were
Power export (GWh) 346.5 333.0 329.7 332.1 induced by fixed and variable operating costs as well as imported elec­
Change (%) − 3.9% − 4.9% − 4.2% tricity. The additional investments related to the additional power de­
Power import (GWh) 71.6 71.6 74.3 71.6
mand nodes in the assessed scenarios did not lead to prominent
Change (%) 0.1% 3.8% 0%
District heat generation by boilers1 132 128 132 132 variations in the annual net profits. However, only the P2CE scenario
(GWh) achieved a positive annual net profit (0.72 M€) with the applied cost
Change (%) − 3.0% 0% 0% assumptions and is hence the most feasible option economically. As the
Biomass consumption (GWh) 107.7 105.8 107.7 107.7 Base (− 0.43 M€), P2EV (− 0.50 M€) and P2H (− 0.27 M€) scenarios led to
Change (%) − 1.8% 0% 0%
negative annual net profits, the value increase derived from processing
Oil consumption (GWh) 24.3 22.3 24.3 10.7
Change (%) − 8.2% 0.0% − 56.0% the selected side and waste material flows outweigh the economic
1
benefits of power exports as well as the income derived from the sold
Includes also an electric boiler in the P2H scenario.

6
K. Kiviranta et al. Applied Energy 279 (2020) 115883

40

30

Total annual costs and income (M€)

Sold electricity (export)
20
Sold electricity (local)
10 Sold district heat
-0.43 0.72 -0.50 -0.27 Sold fuels
0
Imported electricity
Annualized investment costs
-10
Fixed OPEX
-20 Variable OPEX
Total
-30

-40
Base P2CE P2EV P2H

Fig. 3. System annual income and cost structure of the simulated scenarios. The dots indicate the annual net profit of the scenarios.

electricity to electric vehicles or the sold district heat from power-to- biodiesel (15.7 GWh; 107€/MWh). As both fish waste fractionation and
heat conversion in the electric boiler. biodiesel production reached their combined maximum potential for
In comparison to the other scenarios, the P2CE scenario derived electricity consumption (0.15 GWh/a), the maximum amount of bio­
more income from the sold fuels which were produced by the selected diesel (15.7 GWh) was produced. The biodiesel and biomethane could
side and waste material streams as seen in Fig. 4. In the Base, P2EV and be used to replace transport fuels on the islands. When compared to the
P2H scenarios, the annual raw biogas yield was 20.8 GWh, resulting in assumed gasoline and diesel demand of 2025 totaling in 223 GWh, the
an income of 1.25 M€ when applying the biogas price of 60 €/MWh. In fuels in the P2CE scenario could theoretically replace 16.1% of the total
addition, in these scenarios an income of 0.45 M€ was derived from fish consumption, with 7.0% from biodiesel and 9.1% from biomethane. In
waste (3000 t) which was sold as animal fur feed with the associated addition, the raw biogas could be used to produce 7.9 GWh of district
price of 150 €/t. In the P2CE scenario, the annual raw biogas yield was heat, assuming a conversion efficiency of 90%, which would contribute
0.8 GWh higher (21.6 GWh) due to the additional biodegradable waste 6.0% of the annual district heat demand. In both the P2EV and P2H
sources added to the system, from which 8.8 GWh were sold as raw scenarios raw biogas could replace 18.7 GWh and 14.2% of district heat
biogas (60€/MWh) for an income of 0.53 M€. The remaining share of demand with same conversion efficiency (90%), however transportation
raw biogas (12.8 GWh) was directed to the biogas upgrading unit where fuels would not be generated.
an additional SNG yield of 7.5 GWh was achieved by converting carbon As fuel production created the largest benefits in the P2CE scenario
dioxide from the raw biogas to biomethane. The pure biomethane (80 in relation to the alternative scenarios, the sensitivity of product prices
€/MWh) of 20.3 GWh resulted in a total income of 1.62 M€. The biogas (biogas, biomethane, fish waste, biodiesel) on the annual yields of both
upgrading process and hence the production of biomethane from the biomethane and biodiesel are further studied. For both fuels, the model
carbon dioxide was however limited by the availability of hydrogen was able to optimize between two options on an hourly basis: whether
from the electrolyzer, which consumed only 13.4 GWh of electricity and biogas was upgraded to biomethane or sold as a raw biogas, and whether
hence did not reach its annual electricity consumption potential (22.8 fish waste was fractionated and converted to biodiesel or sold as animal
GWh). Consequently, the hydrogen output was not sufficient to convert fur feed instead of the conversion. First, the impact of both raw biogas
all the carbon dioxide from raw biogas into biomethane. and biomethane prices on the annual biomethane yield was studied.
In addition to biomethane, an additional income in the P2CE sce­ Second, the effect of the fish waste and biodiesel price on the annual
nario (1.7 M€) was derived from fish waste which was converted into biodiesel yield was evaluated.

5.0
Annual costs and income of fuels (M€)

4.0

3.0 Income, biodiesel

Income, fish waste

2.0 Income, biomethane

Income, biogas
1.0 Cost, biodiesel
Cost, biomethane
0.0

-1.0
Base P2CE P2EV P2H

Fig. 4. Annual income and cost structure of fuel production in the simulated scenarios.

7
K. Kiviranta et al. Applied Energy 279 (2020) 115883

The prices of the assessed products were varied in steps of 10% to waste is almost tripled during the upgrading process. For example, the
determine the break-even prices between the options in Fig. 5. As seen profit for one ton of fish waste as animal fur feed was 150 € whereas the
from the figure, when decreasing the price of biogas (a), biogas profit for biodiesel processed from it was roughly 410 €. The value in­
upgrading becomes a more favorable option but the yield of biomethane crease is mainly explained by the high expected fat content of the fish,
from the biogas upgrading unit only marginally increases. On the other which directly correlates with the biodiesel yield.
hand, if the price of biogas is increased by 40%, the value of biogas (84
€/MWh) exceeds the value of biomethane (80 €/MWh). However, the 4. Conclusion
yield of biomethane is only halved. As biogas upgrading creates income
also from sold district heat due to hydrogen production, the option is still This study evaluated whether circular economy could increase the
more feasible than the combination of selling raw biogas and exporting value of a renewable energy system in comparison to more conventional
the electricity during power production peaks. When decreasing the alternatives for intermittent power supply utilization by applying a
value of biomethane (b) by 10% in the model, the yield of biomethane combined energy system and material flow optimization. The Åland
begins to decrease linearly until almost no biomethane is produced at Islands were used as a case platform as significant expansions in the local
− 30% of the biomethane value (56 €/MWh) which is below the value of variable renewable energy capacity in the near future will result in high
raw biogas. When the value of biomethane is increased, the yield of deviations in the local power supply. Instead of continuing to export the
biomethane increases only marginally, showing a similar behavior as production exceeding the local demand, part of the electricity could be
when reducing the price of biogas in (a). As local power generation is not utilized to power electric vehicles to generate district heat or to power
available at times, the annual biomethane yield cannot possibly reach its local circular economy processes which aim to upgrade side and waste
theoretical potential (34.3 GWh). The potential is also limited due to material streams into higher-value products. If upgrading the side and
modelling assumptions: the available raw biogas had to be fully waste materials into higher-value products with the intermittent vari­
upgraded or not upgraded at all on an hourly level. Therefore, raw able renewable energy supply would bring financial benefits to a com­
biogas could not be partially upgraded during hours with only a small pany through increased material value, the company itself would have
amount of electricity available for hydrogen production. The price incentives to invest in renewable energy. This is because the circular
variations in fish waste (c) and biodiesel (d) did not have impact on the material upgrading processes would increase the value of the renewable
annual biodiesel yield. With the given cost assumptions, biodiesel pro­ energy investments. As a result, circular economy would act as a catalyst
duction was found to be always more profitable because the value of fish for energy transition. In order to evaluate the hypothesis, a model of the

25 25
Annual biomethane production (GWh)

Annual biomethane production (GWh)

20 20

15 15

10 10

5 5

0 0
-50 -40 -30 -20 -10 0 10 20 30 40 50 -50 -40 -30 -20 -10 0 10 20 30 40 50
Change in the price of biogas (%) Change in the price of biomethane (%)

18 18
Annual biodiesel production (GWh)

Annual biodiesel production (GWh)

16 16

14 14

12 12

10 10

8 8

6 6

4 4

2 2

0 0
-50 -40 -30 -20 -10 0 10 20 30 40 50 -50 -40 -30 -20 -10 0 10 20 30 40 50
Change in the price of fish waste (%) Change in the price of biodiesel (%)
Fig. 5. Sensitivity analysis for the annual production of biomethane as a function of (a) price of biogas and (b) price of biomethane, and for the annual production of
biodiesel as a function of (c) price of fish waste and (d) price of biodiesel in the power-to-circular economy scenario.

8
K. Kiviranta et al. Applied Energy 279 (2020) 115883

circular ecosystem was created for the Åland Islands on the basis of Table A1
gathered material flow information based on official statistics, research Symbols used in the model equations.
literature and publicly available data from companies located in the Notation Description
Åland Islands.
p Parameter
The analysis of the study shows that increased material value via the v Variable
utilization of local power production peaks in regional circular processes t Hour index
can increase the economic profitability of the energy system. According T Set of hours (1 … 8760)
to the results, the energy system combined with circular processes (0.72 u Unit
U Set of units
M€) was more economically feasible than the alternative of exporting
the production exceeding the local demand (− 0.43 M€), partly electri­
fying the transportation sector (− 0.5 M€) and generating district heat circular ecosystem would require changes to the current system, and
(− 0.27 M€) in terms of system’s annual net profit. However, the po­ therefore cooperation between local companies. As for the Åland
tential to increase the value of the system was constrained by two fac­ Islands, the fish waste sources are currently separated in different lo­
tors: the amount of electricity that can be integrated, and the maximum cations and would require transporting to a centralized processing
capacity of the circular processes. In this case, the benefits were limited location. Similarly, all the biodegradable waste streams generated in the
by the local availability of biodegradable feedstock for biogas and bio­ system should be collected to be processed by one centralized biogas
diesel production, as well as the variable renewable energy for hydrogen unit. If such a unit were not implemented, the income from upgrading
production required for the upgrading of biogas. Furthermore, when the biogas would decrease, and consequently so would the feasibility of
creating a circular ecosystem, closed-loop material cycles were found to the circular economy scenario. The other alternative scenarios involved
already exist in the Åland Islands. For example, several companies fewer limitations regarding the required investments: in reality, the
already converted their own biodegradable side products into biogas system could be enabled by private consumers investing in electric ve­
and further into energy, and the side products from the wood industry hicles or a district heating company. Therefore, the study highlights how
were already being used for local district heating. Therefore, a circular circular economy is able to create economic benefits but also faces non-
ecosystem could not be created by directly connecting the material and economic barriers before the benefits can be realized from a conceptual
energy streams of different processes but would require the integration level to reality.
of technologies which are novel to the system.
Synergies of connecting energy system with circular economy should CRediT authorship contribution statement
be studied on a regional level as circular processes are always local.
Although the study shows circular economy to be an economically Kirsikka Kiviranta: Methodology, Investigation, Writing - original
attractive option, both the effect of the assumptions and the practical draft, Visualization. Tomi Thomasson: Methodology, Software, Vali­
requirements should be considered. The study assumed a value increase dation, Writing - review & editing. Jonne Hirvonen: Conceptualization,
and therefore a higher price for the upgraded circular economy products Investigation. Matti Tähtinen: Conceptualization, Funding acquisition,
(upgraded biogas and biodiesel) compared to the unprocessed alterna­ Supervision.
tives (raw biogas and fish waste). The product values were considered
realistic for the studied system, as they were estimates based on national
values. However, some uncertainty is present related to the investment Declaration of Competing Interest
costs for the circular economy processes as well as the material flow
volumes. While investment costs were available for similar systems, for The authors declare that they have no known competing financial
example fractionation or biodiesel production, the values were scaled interests or personal relationships that could have appeared to influence
based on the production capacities of the studied system. As the data for the work reported in this paper.
the material flow volumes was aggregated from multiple sources, the
assumed volumes do not necessarily reflect the real situation. Further­ Acknowledgements
more, most of the material flows were assumed to be continuous and
equally distributed. If the material flows were available in batches with The research for this study was performed as part of VTT’s research
varied volumes, less variable renewable energy could potentially be program “Circular Economy Concepts”. The authors are grateful for the
used in the circular economy processes. On practical level, the created financial support.

Appendix A. Model description

The optimization model operates at an hourly resolution with full foresight throughout the year. The symbols used in the model equations are
introduced in Table A.1.

Objective function

The objective function (Eq. (A.1)) of the model includes the cost and income terms relevant for the studied system, which consist of the income
from sold electricity (Eq. (A.2)), the income from sold district heat (Eq. (A.3)), the income from exported electricity (Eq. (A.4)), fuel costs (Eq. (A.5)),
other variable operating costs (Eq. (A.6)), emission allowance costs depending on the fossil fuel share in the unit’s fuel mix (Eq. (A.7)), and costs from
imported electricity. In addition, the objective function includes income terms for sold biogas and biodiesel as well as cost terms for biodiesel pro­
duction, which are introduced later.
( )
∑ vin,el + vin,th + vin,export + vin,biogas + vin,biodiesel
obj
v = t t t t t
(A.1)
t∈T − vcost,fuel
t − vcost,OM,var
t − vcost,emission
t − vcost,import
t − vcost,biodiesel
t

vin,el
t = (psolar
t + vwind,local
t )⋅pprice,el
t + vel,vehicles
t ⋅pprice,el
t (A.2)

9
K. Kiviranta et al. Applied Energy 279 (2020) 115883

∑
vin,th
t = vth
u,t ⋅p
price,th
+ (vth,heat
t + vth,circular
t )⋅pprice,th (A.3)
u

(A.4)
price,el
vin,export
t = vwind,export
t ⋅pt
∑
vcost,fuel
t = pfuel,nom
u ⋅vload price,fuel
u,t ⋅pu (A.5)
u

∑
vcost,OM,var
t = pfuel,nom
u ⋅vload cost,OM,var
u,t ⋅pu (A.6)
u

∑
vcost,emission
t = pfuel,nom
u ⋅vload
u,t ⋅p
price,CO2 share,fossil
⋅p (A.7)
u

(A.8)
price,el
vcost,import
t = vel,import
t ⋅pt

In addition to the variable terms, fixed costs are calculated to determine the system’s annual net profit. Eq. (A.9) presents the calculation of the
fixed operating costs for the power generation and circular economy based on the unit nominal capacity. The annualized investment cost for the unit
and process investments is calculated in Eq. (A.11) using the weighted average cost of capital, investment lifetime, and the investment present value
(Eq. (A.10)).
pcost,OM,fix = pnom cost,OM,fix
u ⋅pu (A.9)

pcost,investment = pnom cost,inv,spec

u ⋅pu (A.10)
[ ]
( ) lifetime ( ) lifetime
ppayment
u,t = pcost,investment ⋅pWACC
u ⋅ 1 + p WACC pu
u / 1 + p WACC pu
u − 1 (A.11)

For all dispatchable units generating heat, the fuel input is converted into heat output based on the load percentage and conversion efficiency, as
shown by Eq. (A.12)
vth fuel,nom load eff,th
u,t = pu ⋅vu,t ⋅pu (A.12)

Variable renewable energy

Depending on the scenario options, the hourly wind generation is constrained equal to the sum of different option variables representing the
utilization options (Eq. (A.13)). In all scenarios, local consumption and export options are available, whereas heat and circular options are added in the
power-to-heat and power-to-circular economy scenarios, respectively.
pwind
t = vwind,local
t + vwind,export
t + vwind,heat
t + vwind,circular
t (A.13)

System balances

Energy balance equations constrain the hourly operation of the power system (Eq. (A.14)) and district heating system (Eq. (A.15)). In the power
system, the hourly demand is constrained equal to the sum of power transmission (imported power and exported wind), wind power used locally, solar
power used locally, and the additional demand caused by the different scenarios. Similarly, the heat demand is constrained equal to the sum of unit
production, and the production or demand of the different scenarios. The possibility to overproduce the demand or curtail the variable renewable
power generation is not included. For the power transmission (Eqs. (A.16) and (A.17)), the hourly transmission is constrained below the maximum
transmission capacity.
pelt = vel,import
t − vwind,export
t + vwind,local
t + psolar
t − vel,circular
t − vel,vehicles
t (A.14)
∑
pth
t = vth th,heat
u,t + vt + vth,circular
t (A.15)
u∈U

vel,import
t ≤ ptrans,max
t (A.16)

vwind,export
t ≤ ptrans,max
t (A.17)

Power-to-electric vehicles

In the power-to-electric vehicles (“P2EV”) scenario, electric vehicles are considered as an additional power demand node that is included in the
power system balance. The daily total charging requirement is determined by Eq. (A.18) using the number of electric vehicles, the average daily road
transportation demand, and the specific power consumption, whereas Eq. (A.19) defines the total hourly maximum charging rate. The parameters are
used with two constraints: Eq. (A.20) is applied in daily windows to ensure that the total daily charging requirement is met, and Eq. (A.21) is used to
limit the hourly charging rate.

10
K. Kiviranta et al. Applied Energy 279 (2020) 115883

pvehicles,demand = pn,vehicles ⋅pvehicles,distance ⋅pvehicles,demand (A.18)

pvehicles,peak = pn,vehicles ⋅pvehicles,charge,max (A.19)

∑
vel,vehicles
t = pel,vehicles,demand (A.20)
t

vel,vehicles
t ≤ pvehicles,peak (A.21)

Power-to-heat

The power-to-heat (“P2H”) scenario adds electric boiler capacity to the energy system. The capacity is considered as a heat generation unit, and its
production (Eq. (A.22)) is therefore included in the district heating system balance equation. Consequently, income is gained from the sold district
heat.
vth,heat
t = vwind,heat
t ⋅peff,boiler,electric
u (A.22)

Power-to-circular economy

The power-to-circular economy (“P2CE”) scenario adds multiple new elements to the energy system, affecting the power system and district
heating system balance equations with new demand and production terms (Eqs. (A.23) and (A.24)). The electricity consumed by the circular processes
is constrained equal to the wind power allocated for the circular processes (Eq. (A.25)).
vel,circular
t = vel,electrolysis
t + vel,biodiesel
t (A.23)

vth,circular
t = vth,electrolysis
t − vth,biodiesel
t (A.24)

vwind,circular
t = vel,circular
t (A.25)

The hydrogen production of the electrolyzer is calculated using Eq. (A.26). Part of the electricity used is turned into process heat (Eq. (A.27)),
which is integrated into the district heating system energy balance.
vH2,electrolysis
t = vel,electrolysis
t ⋅peff,electrolysis (A.26)

vth,electrolysis
t = vel,electrolysis
t − vH2,electrolysis
t (A.27)

The produced hydrogen is stored in short-term storage, for which the hourly capacity is constrained between the minimum and maximum capacity
(Eq. (A.28)). The hydrogen discharged from the storage can be presented as the change in capacity in consecutive hours (Eq. (A.29)). The hydrogen
storage is initialized with half of the maximum capacity (Eq. (A.30)) and is expected to return to the initial state at the end of the time horizon (Eq.
(A.31)).
0 ≤ vH2,storage,cap
t ≤ pH2,storage,cap,max (A.28)

vH2,storage,discharge
t = vH2,storage,cap
t+1 − vH2,storage,cap
t (A.29)

vH2,storage,cap
1 = 0.5⋅pH2,storage,cap,max (A.30)

vH2,storage,cap
1 = vH2,storage,cap
T (A.31)
The hydrogen storage is connected to the upgrading of the biogas. In Eqs. (A.32) and (A.33), the hourly raw biogas yield is used to determine the
hourly volumes of methane and carbon dioxide, respectively. This purified methane can be directly sold whereas the carbon dioxide fraction is
upgraded to biomethane using hydrogen. The Sabatier reaction (Eq. (A.34)), along with molar volumes and molar masses of carbon dioxide and
hydrogen are used to determine the amount of hydrogen required for the upgrade as well as the resulting amount of synthetic natural gas. After
upgrading, the total amount of biomethane therefore consists of two components; purified methane and upgraded synthetic natural gas (Eq. (A.35)).
pCH4,purified
t = pbiogas
t ⋅pfraction,CH4 (A.32)

pCO2,purified
t = pbiogas
t ⋅(1 − pfraction,CH4 ) (A.33)

CO2 + 4H2 →CH4 + 2H2 O (A.34)

pCH4,total
t = pCH4,purified
t + pCH4,upgraded
t (A.35)

The upgrading of the biogas is controlled with a binary variable (Eq. (A.36)). A constraint is applied to ensure that the hydrogen required for the
upgrading, solved from the Sabatier reaction, is equal to the amount extracted from the hydrogen storage (Eq. (A.37)).
vH2,upgrading
t = pH2,required
t ⋅vbiogas,upgraded
t (A.36)

vH2,upgrading
t = vH2,storage,discharge
t (A.37)

11
K. Kiviranta et al. Applied Energy 279 (2020) 115883

Eq. (A.38) shows the income function for biogas production, in which a binary variable is also used to control the income from upgraded biogas. In
scenarios other than power-to-circular economy, the option for the upgrading of biogas is excluded and the income function is simplified to Eq. (A.39),
in which the amount and price of raw biogas is applied.
( ) CH4,purified price,biogas
vin,biogas
t = 1 − vbiogas,upgraded
t ⋅pt ⋅p + vbiogas,upgraded
t ⋅(pCH4,purified
t + pCH4,upgraded
t )⋅pprice,biogas,upgraded (A.38)

vin,biogas
t = pbiogas
t ⋅pprice,biogas (A.39)

For biodiesel production, the fish waste is divided into two fractions based on the hourly fish waste amount using a binary variable.
vwaste,feed
t = pwaste,amount
t ⋅(1 − vwaste,upgraded
t ) (A.40)

vwaste,biodiesel
t = pwaste,amount
t ⋅vwaste,upgraded
t (A.41)

For the waste fraction used as fur animal feed (Eq. (A.40)), the income can be directly calculated (Eq. (A.42)).
vin,waste,feed
t = vwaste,feed
t ⋅pprice,waste,feed (A.42)

For the waste fraction to be upgraded into biodiesel (Eq. (A.41)), the amounts of processed waste oil (Eq. (A.43)), biodiesel (Eq. (A.44)) and
glycerin (Eq. (A.45)) fractions are calculated based on defined yield percentages.
voil,amount
t = vwaste,biodiesel
t ⋅pfraction,oil (A.43)

vbiodiesel,amount
t = voil,amount
t ⋅pfraction,biodiesel (A.44)

vglyserin,amount
t = vbiodiesel,amount
t ⋅pfraction,glyserin
t (A.45)

For the upgrading steps, the associated income and cost functions are defined in Eqs. (A.46) and (A.47). The income function (Eq. (A.48)) consists
of income from sold biodiesel and glycerin, whereas the cost function (Eq. (A.49)) sums the costs from the required formic acid and methanol.
vin,biodiesel
t = vbiodiesel,amount
t ⋅pprice,biodiesel + vglyserin,amount
t ⋅pprice,glyserin + vin,waste,feed
t (A.46)

vcost,biodiesel
t = voil,amount
t ⋅preq,acid
t ⋅pprice,acid + vbiodiesel,amount
t ⋅preq,methanol
t ⋅pprice,methanol (A.47)

Finally, the power and heat demand of the upgrading steps are calculated based on the specific energy consumption in Eqs. (A.48) and (A.49).
vel,biodiesel
t = voil,amount
t ⋅pel,req,oil + vbiodiesel,amount
t ⋅pel,req,biodiesel (A.48)

vth,biodiesel
t = voil,amount
t ⋅pth,req,oil + vbiodiesel,amount
t ⋅pth,req,biodiesel (A.49)

References [13] Meschede H, Holzapfel P, Kadelbach F, Hesselbach J. Classification of global island

regarding the opportunity of using RES. Applied Energy 2016;175:251–8. https://
doi.org/10.1016/j.apenergy.2016.05.018.
[1] United Nations. Paris Agreement; 2015. [Online]. Available: https://unfccc.int/site
[14] Child M, Nordling A, Breyer C. Scenarios for a sustainable energy system in the
s/default/files/english_paris_agreement.pdf. [accessed: 13-Aug-2019].
Åland Islands in 2030. Energy Conversion and Management 2017;137:49–60.
[2] Frankfurt School-UNEP Centre/BNEF. Global trends in renewable energy
https://doi.org/10.1016/j.enconman.2017.01.039.
investment 2018; 2018. [Online]. Available: https://www.fs-unep-centre.org/
[15] Cross S, Padfield D, Ant-Wuorinen R, King P, Syri S. Benchmarking island power
[accessed: 13-Aug-2019].
systems: Results, challenges, and solutions for long term sustainability. Renewable
[3] IRENA. Renewable power generation costs in 2018. Abu Dhabi: International
and Sustainable Energy Reviews 2017;80:1269–91. https://doi.org/10.1016/j.
Renewable Energy Agency; 2019.
rser.2017.05.126.
[4] IEA. Renewable capacity growth worldwide stalled in 2018 after two decades of
[16] Meschede H, Child M, Breyer C. Assessment of sustainable energy system
strong e; 2019. [Online]. Available: https://www.iea.org/news/renewable-capa
configuration for a small Canary island in 2030. Energy Conversion and
city-growth-worldwide-stalled-in-2018-after-two-decades-of-strong-expansion
Management 2018;165:363–72. https://doi.org/10.1016/j.
[accessed: 13-Aug-2019].
enconman.2018.03.061.
[5] IEA. Global Energy & CO2 Status Report; 2019. [Online]. https://www.iea.org/re
[17] Gils HC, Simon S. Carbon neutral archipelago – 100% renewable energy supply for
ports/global-energy-co2-status-report-2019 [accessed: 13-Aug-2019].
the Canary Islands. Applied Energy 2017;188:342–55. https://doi.org/10.1016/j.
[6] IRENA. Global Energy Transformation: A roadmap to 2050. Abu Dhabi:
apenergy.2016.12.023.
International Renewable Energy Agency; 2018.
[18] Al Katsaprakakis D, Thomsen B, Dakanali I, Tzirakis K. Faroe Islands: Towards
[7] Lantto R, Järnefelt V, Tähtinen M, Jääskeläinen A-S, Laine-Ylijoki J, Oasmaa A,
100% R.E.S. penetration. Renewable Energy 2019;135:473–84. https://doi.org/
et al. Commentary going beyond a circular economy: a vision of a sustainable
10.1016/j.renene.2018.12.042.
economy in which material, value and information are integrated and circulate
[19] Garcia Latorre FJ, Quintana JJ, de la Nuez I. Technical and economic evaluation of
together. Industrial Biotechnology 2019;15(1):12–9. https://doi.org/10.1089/
the integration of a wind-hydro system in El Hierro island. Renewable Energy
ind.2019.29156.rla.
2019;134:186–93. https://doi.org/10.1016/j.renene.2018.11.047.
[8] IRENA. Power system flexibility for the energy transition, Part 1: Overview for
[20] Notton G, Nivet ML, Zafirakis D, Motte F, Voyant C, Fouilloy A. Tilos, the first
policy makers. Abu Dhabi: International Renewable Energy Agency; 2018.
autonomous renewable green island in Mediterranean: A Horizon 2020 project. In:
[9] Cochran J, Miller M, Zinaman O, Milligan M, Arent D, Palmintier B, et al.
15th International Conference on Electrical Machines. Sofia, Bulgaria: Drives and
Flexibility in 21st century power systems. NREL 2014.
Power Systems (ELMA 2017); 2017. p. 102–5. https://doi.org/10.1109/
[10] Paiho S, Saastamoinen H, Hakkarainen E, Similä L, Pasonen R, Ikäheimo J, et al.
ELMA.2017.7955410.
Increasing flexibility of Finnish energy systems—A review of potential technologies
[21] Parkkari J. Ahvenanmaasta Suomen seuraava vientimenestys? Maakunta siirtyy
and means. Sustainable Cities and Society Nov. 2018;43:509–23. https://doi.org/
uusiutuvaan energiaan, ja malli halutaan myydä Kiinaan ja Intiaan [Is Åland the
10.1016/j.scs.2018.09.015.
Finland’s next export success story? The region is transitioning to renewable
[11] Lund PD, Lindgren J, Mikkola J, Salpakari J. Review of energy system flexibility
energy and wants to sell the model to China and India]; YLE, 2018. [Online].
measures to enable high levels of variable renewable electricity. Renewable and
Available: https://yle.fi/uutiset/3-10127907 [accessed: 22-Oct-2018].
Sustainable Energy Reviews May. 2015;45:785–807. https://doi.org/10.1016/j.
[22] Thomasson T, Kiviranta K, Hirvonen J, Tähtinen M. Assessing scenarios for fully
rser.2015.01.057.
renewable Åland Islands. Manuscript Submitted for publication 2020.
[12] Lantto (Ed) R, Järnefelt (Ed) V, Tähtinen M. Going beyond a circular economy. VTT
[23] Allwinds Ab. De åländska vindkraftverken [The wind power plants of Åland];
Technical Research Centre of Finland. VTT Visions 2018. https://www.vtt.fi/inf
2018. [Online]. Available: https://www.allwinds.ax/de-alandska-vindkraft
/pdf/visions/2018/Going_beyond_a_circular_economy.pdf.
verken/ [accessed: 05-Feb-2018].

12
K. Kiviranta et al. Applied Energy 279 (2020) 115883

[24] Allwinds Ab. The Wind Power of Åland Islands; 2018. [Online]. Available: https:// ub.ax/PXWeb/pxweb/en/Statistik/Statistik__MI/MI003.px/table/tableViewLayout
www.allwinds.ax/wp-content/uploads/2018/01/Allwinds_Smart-Energy-Island. 2/?rxid=5fa57277-5d80-4c56-942f-20825742f80c [accessed: 27-Jun-2019].
pdf. [accessed: 05-Feb-2018]. [50] B. Bergman. Fiskrens, en outnyttjad resurs [Fish waste, an untapped resource];
[25] Child M, Nordling A, Breyer C. The impacts of high V2G participation in a 100 % 2017. [Online]. Available: https://fiskodlarna.ax/wp-content/uploads/2020/02/
renewable Åland energy system. Energies 2018;11. https://doi.org/10.3390/ Bilaga-7.-Fiskrens-en-outnyttjad-resurs.pdf [Accessed: 27 -Jun-2019].
en11092206. [51] Ronde H, Ranne A, Pursiheimo E. Integrated renewable energy solutions for
[26] Pääkkönen A, Joronen T. Revisiting the feasibility of biomass-fueled CHP in future seafood processing stations. In: Proceedings of the International Conference on
energy systems – Case study of the Åland Islands. Energy Conversion and Energy and Sustainable Development: Issues and Strategies (ESD 2010): 2010 June
Management May 2019;188:66–75. https://doi.org/10.1016/j. 2-4, Chiang Mai, Thailand. doi:10.1109/ESD.2010.5598777; 2010.
enconman.2019.03.057. [52] Flexens – Smart Energy Åland. Bioenergi och avfallsbaserad energiproduktion
[27] Hart WE, et al. Pyomo: modelling and solving mathematical programs in Python. [Bioenergy and waste-based energy production]; 2020. [Online]. Available:
Mathematical Programming Computation 2011;3(219). https://doi.org/10.1007/ https://smartenergy.ax/biogas/.
s12532-011-0026-8. [53] Allerborg M, Bergström J, Englöf J, Holroyd J, Hultin C, Reuter O. Rötning av
[28] IBM. IBM CPLEX Optimizer; 2020. [Online]. Available: https://www.ibm. bioavfall på Åland [Digestion of biowaste in Åland]. KTH Stockholm; 2015.
com/analytics/cplex-optimizer. [54] Stallhagen. Årsberättelse 2017 [Annual report 2017]; 2017. [Online]. Available:
[29] Kraftnät Åland Ab. Kraftnät Åland Årsberättelse 2007–2018 [Annual report of http://www.stallhagen.com/content/uploads/2018/05/arsberattelse17_webb.pdf
Kraftnät Åland 2007-2018]; 2019. [Online]. Available: http://www.kraftnat.aland. [accessed: 27-Jun-2019].
fi/text.con?iPage=11&m=21 [accessed: 27-Jun-2019]. [55] Niemi P. Enzymatic fractionation of brewer’s spent grain and bioconversion of
[30] Finnish Energy. District heating statistics; 2018. [Online]. Available: https://ene lignin-rich fractions in a colon model in vitro; 2016. [Dissertiation].
rgia.fi/en/news_and_publications/publications/district_heating_statistics. [56] Kärki J, Kouri S, Hurskainen M, Tsupari E, Alakangas E, Bajamundi C. Integrated
html#material-view [accessed: 15-Oct-2018]. utilization pathways for biogenic CO2 in biomass driven industry sectors; 2017.
[31] Mariehamns Energi Ab. Årsberättelse 2017 [Annual report 2017]; 2018. [Online]. [Online]. Available: https://projectsites.vtt.fi/sites/BioCO2/www.vtt.fi/sites/BioC
Available: https://issuu.com/straxkommunikation/docs/arsberattelse_2017. O2/PublishingImages/tiedotteet/5BO.4.4_presentation.pdf [accessed: 27-Jun-
[accessed: 05-Feb-2018]. 2019].
[32] Statistics and Research Åland (ÅSUB). Supply of some of the common oil products [57] Jentsch M, Trost T, Sterner M. Optimal use of power-to-gas energy storage systems
by the oil companies in Åland 1982-2018 by year and product; 2019. [Online]. in an 85% renewable energy scenario. Energy Procedia 2014;46:254–61. https://
Available: https://pxweb.asub.ax/PXWeb/pxweb/en/Statistik/Statistik__MI/?rx doi.org/10.1016/j.egypro.2014.01.180.
id=4f7d9b63-2d7c-4db9-aa9c-603a83bd8d03 [accessed: 20-Nov-2019]. [58] Lecker B, Illi L, Lemmer A, Oechsner H. Biological hydrogen methanation – A
[33] Finnish Meteorological Institute. Havaintojen lataus [Download the observations]; review. Bioresource Technology 2017;245(Part A):1220–8. https://doi.org/
2018. [Online]. Available: https://ilmatieteenlaitos.fi/havaintojen-lataus#!/ 10.1016/j.biortech.2017.08.176.
[accessed: 08-Nov-2018]. [59] Tsupari E, Kärki J, Vakkilainen E. Economic feasibility of power-to-gas integrated
[34] Enercon GmbH. E-44; 2018. [Online]. Available: https://www.enercon.de/en/pr with biomass fired CHP plant. Journal of Energy Storage 2016;5:62–9. https://doi.
oducts/ep-1/e-44/ [accessed: 08-Nov-2018]. org/10.1016/j.est.2015.11.010.
[35] Enercon GmbH. E-70; 2018. [Online]. Available: https://www.enercon.de/en/pr [60] S. Kouri, M. Hurskainen, J. Kärki, E. Tsupari, E. Alakangas, and C. Bajamundi.
oducts/ep-2/e-70/ [accessed: 08-Nov-2018]. Integrated utilisation pathways for biogenic carbon dioxide in biomass driven
[36] Enercon GmbH. E-101; 2018. [Online]. Available: https://www.enercon.de/en/p industry sectors. Proceedings of the 25th European Biomass Conference and
roducts/ep-3/e-101/ [accessed: 08-Nov-2018]. Exhibition (EUBCE 2017): 2017 June 12-15, Stockholm, Sweden. 2017.
[37] Government of Åland. Energi- och klimatstrategi för Åland till år 2030 [Energy and [61] ÅSUB. Statistisk årsbok för Åland 2018 [Statistical Yearbook of Åland 2018].
Climate Strategy of Åland for 2030]; 2017. [Online]. Available: https://www. Mariehamn; 2018. [Online]. Available: https://www.asub.ax/sites/www.asub.
regeringen.ax/sites/www.regeringen.ax/files/attachments/page/lr[KK1]_energi_k ax/files/attachments/page/statistisk_arsbok_for_aland_2018_1.pdf [accessed: 27-
limatstrat_2030.pdf [accessed: 08-Nov-2018]. Jun-2019].
[38] Nordpool. Historical Market Data; 2018. [Online]. Available: https://www.nor [62] Ruska M, Kiviluoma J, Koreneff G. Sähköautojen laajan käyttöönoton skenaarioita
dpoolgroup.com/historical-market-data/ [accessed: 15-Aug-2019]. ja vaikutuksia sähköjärjestelmään [Scenarios of large-scale deployment of electric
[39] CLIC Innovation. Final joint report of the projects FLEXe DEMO and CEMBioFlex; vehicles and their power system impacts]. VTT Technical Research Centre of
2019. [Online]. Available: https://flexens.com/wp-content/uploads/2019/03/Fin Finland; 2010. [Online]. Available: https://www.vttresearch.com/sites/default/fil
al_Report_FLEXe_demo_and_CEMBioFlex.pdf [accessed: 15-Aug-2019]. es/pdf/workingpapers/2010/W155.pdf [accessed: 27-Jun-2019].
[40] Pöyry. Kraftnät Åland, Finland; 2016. [Online]. Available: https://www.poyry. [63] Finnish Transport Infrastructure Agency. Henkilöliikennetutkimus 2016
com/sites/default/files/media/related_material/kraftnat_aland_case_story_a4_new. [Passenger Transport Survey 2016]; 2018. [Online]. Available: https://julkaisut.
pdf [accessed: 05-Feb-2018]. vayla.fi/pdf8/lti_2018-01_henkiloliikennetutkimus_2016_web.pdf [accessed: 27-
[41] Ålands Skogsindustrier Ab. Verksamhetsberättelse 2017 [Annual report 2017]; Jun-2019].
2018. [Online]. Available: http://www.skogsindustrier.ax/sites/www.skogsindust [64] Paakkinen M, Pihlatie M, Peltola V, Pylsy P. Sähköautojen kotilataaminen:
rier.ax/files/attachments/verksamhetsberattelse_2017.pdf [accessed: 13-Dec- GASELLI-väliraportti 1. VTT Tutkimusraportti, Nro VTT-R-02416-18 [Home
2018]. charging of electric vehicles: First interim report of the GASELLI project]. VTT
[42] Ålandsmejeritet. Energisamarbete [Energy cooperation]; 2018. [Online]. Technical Research Centre of Finland; 2018.
Available: https://www.aca.ax/sites/www.aca.ax/files/upload/gastronomen_we [65] Ikäheimo J, Kiviluoma J, Weiss R, Holttinen H. Power-to-ammonia in future North
bb.pdf [accessed: 13-Dec-2018]. European 100 % renewable power and heat system. International Journal of
[43] Avsättning av slutprodukterna från en åländsk biogasanläggning [Disposal of the Hydrogen Energy 2018;43(36):17295–308. https://doi.org/10.1016/j.
end products from a biogas plant in Åland] Renhållningsmyndigheten Ålands ijhydene.2018.06.121.
Miljöservice KF, 2018. [66] ENERFISH Consortium. Integrated renewable energy solutions for seafood
[44] Mariehamns Stad. Lotsbroverket; 2010. [Online]. Available: https://www.marieh processing stations; 2012. [Online]. Available: https://www.shetlandrenewables.
amn.ax/globalassets/dokument/06_boende_miljo/boende/lotsbroverket_infobro com/assets/files/WP6%20Enerfish%20Report%20D13%20Final%20Report.pd
schyr.pdf [accessed: 24-Jun-2019]. [accessed: 27-Jun-2019].
[45] Fifax. Fifax Ab; 2019. [Online]. Available: https://fifax.ax/en/om-oss/ [accessed: [67] Danish Energy Agency and Energinet. Technology Data - Energy Plants for
24-Jun-2019]. Electricity and District heating generation. Danish Energy Agency; 2020.
[46] VTT. Puhtaan veden ja ravinteiden talteenotto saariston, veneiden ja [68] Statistics Finland. Statistics: Energy prices; 2018. [Online]. Available: http://www.
kalajalostamon jätevesistä etenee koevaiheeseen [Recovery of clean water and stat.fi/til/ehi/2018/02/ehi_2018_02_2018-09-12_tie_001_en.html [accessed: 23-
nutrients from the wastewater of the archipelago, boats and fish processing plant is Nov-2018].
progressing to piloting]; 2019. [Online]. Available: https://www.vtt.fi/medialle/ [69] Finnish Energy. District heating in Finland 2017; 2018. [Online]. Available:
uutiset/puhtaan-veden-ja-ravinteiden-talteenotto-saariston-veneiden-ja-kalajalo https://energia.fi/files/2948/District_heating_in_Finland_2017.pdf [accessed: 23-
stamon-jätevesistä-etenee-koevaiheeseen [accessed: 25-Jun-2019]. Nov-2018].
[47] B. Bergman. Mervärde ur fiskrens-Projektrapport [Added value from fish waste - [70] Statistics Finland. Sähkön ja lämmön tuotanto 2017 [Power and heat production
project report]; 2017. [Online]. Available: https://www.ha.ax/uploads/2018/0 2017]; 2018. [Online]. Available: https://www.stat.fi/til/salatuo/2017/sal
4/bergman_mervarde-ur-fiskrens_2017.pdf [accessed:24-Jun-2019]. atuo_2017_2018-11-01_tie_001_fi.html [accessed: 13-Apr-2019].
[48] ÅSUB. Avfallsstatistik 2016 [Waste statistics 2016]; 2017. [Online]. Available: htt [71] EEX. European Emission Allowances (EUA); 2019. [Online]. Available:
ps://www.asub.ax/sites/www.asub.ax/files/statistics/avfall16.pdf [accessed: 24- https://www.eex.com/en/market-data/environmental-markets/auction-
Jun-2019]. market/european-emission-allowances-auction#!/2019/12/10 [accessed: 10-Dec-
[49] Statistics and Research Åland (ÅSUB). Generated waste in Åland by waste type, 2019].
year and hazardous/non-hazardous; 2017. [Online]. Available: https://pxweb.as

13