Life cycle assessment
This study presents an LCA of heat supply scenarios for the replacement of an existing gas-fired boiler supplying heat for three buildings located in Dublin, Ireland. The three replacement systems considered are a waste-heat fed heat pump district heating system, a biomass CHP plant district heating system, and an individual gas boiler system. The LCA was carried out in accordance with ISO14040 standards in four steps i.e., goal and scope definition, inventory analysis, impact assessment and interpretation. The LCA was performed using the 8.2.0.55 version of Gabi which is an LCA accounting tool developed by Thinkstep (Thinkstep 2018).
Scope and Functional unit
The scope of the current study was to evaluate the life cycle environmental impacts of waste heat fed heat pump district heating system (WHP-DH) and biomass CHP plant district heating system (BCHP-DH) and compare with an individual gas boiler system (GB). The system boundary was considered as ‘cradle to gate’ that includes infrastructure facility (equipment, construction work and buildings related to each system), equipment raw material extraction, fuel extraction, electricity generation, and distribution and transportation of raw materials.
As the replacement heating system in the SDCC buildings was only in the planning stage this LCA is prospective using data provided by Dublin’s Energy Agency (Codema 2015) and published LCA studies. Moreover, the study area has heat densities two to three times above 150TJ/km2 which is considered highly feasible for district heating systems from a Danish perspective (Euroheat and Power 2013). As stated previously, the boundary “cradle to gate’ would follow the study carried out by Nitkiewicz and Sekret (2014) which states a 20-year lifecycle and excludes disposal due to probable advances within the timespan considered in the LCA. 20 years was chosen as it is the lifespan of most of the equipment, but it should be noted that the pipeline’s lifespan is much longer. Materials to make the equipment specifically for the system, such as the heat pump, were included. Avoided use of other fuels due to the energy generated within the system were also not included. A system diagram showing the processes within the cradle-to-gate the system boundary is shown in Fig. 1.
In the heat pump DH system (WHP-DH), the boundary starts at the collection of the waste heat and includes all necessary steps, along with equipment, needed to process this heat. The product is the heat produced. The BCHP-DH system includes the generation of the fuel from the forest nursery phase comprising all the infrastructure necessary to deliver the heat to the required buildings. Waste disposal and the machinery needed for generating fuel or construction in all phases were excluded.
The main function of the studied system was to produce heat for the SDCC buildings, so the functional unit is 1 MWh of heat produced. In a CHP plant, both electricity and heat are useful outputs, so electricity was allocated a proportion to the environmental impact. In the absence of Irish case study data this allocation was based on an average Irish CHP efficiency of 86% (54% Heat and 32% electricity production) (SEAI 2015).
System description
Data centres are computer warehouses that store data, with high energy demands and large amounts of heat generated. In 2016, total data centre energy use made up 4.8% of Ireland’s total electricity demand (Bitpower 2017). Lu et al. (2011) used real production data from Finland to calculate that waste heat could be captured from 97% of the total power consumed in the data centre creating large sources of potential heat. The use of waste heat from data centres to power a district heating system via a heat pump has been demonstrated in Mäntsälä, Finland (EHPA 2019).
In the first system (WHP-DH), waste heat from the data centre was harnessed through a collector coil with exhaust air heating water of 20–35°C. This warm water gets heated in the energy centre by an ammonia heat pump with a seasonal performance factor of 3.6, to 70°C on the primary side. Average Irish grid electricity was assumed with a carbon intensity of 428 kg CO2/MWh (SEAI 2018). The hot water was transported 701 metres through a heat exchanger and distributed to the customer. The end user buildings had the existing gas boiler replaced by a pump, control valves and a meter situated in the heat exchanger substation. The water was pumped to the council building with the existing radiator system being maintained. The total installed pipeline was 2,142 metres long, consisting of a 5MW gas boiler to top up the temperature of the DH system in winter and act as a back-up. The gas boiler provides 20% of the heat with the heat pump providing the remainder.
The second system consisting of a 5MW Biomass CHP district heating system (BCHP-DH) was built on the Council brownfield site to meet the base load while the existing gas boilers were replaced to meet the peak demand. A feasibility study on the potential DH system in Tallaght was found to have a payback period of 15 years (Gartland 2014). The CHP produces temperatures of 75°C and has an efficiency of 86% (54% heat and 32% electricity production) (SEAI 2015). A pipeline of 137 metres connects the CHP to the Council building. The CHP plant transports hot water to a heat exchanger substation after which it is distributed to the end user. The impacts of biomass production and processing was based on the biomass supply chains in Murphy et al. (2014) and Murphy et al. (2016). Biomass consisted of wood chips sourced from the Laois, Ireland, area.
The third system consists of maintaining the existing individual gas boilers in each of the Council buildings. These boilers were installed 18 years ago so require replacement. The LCA examines the replacement of these boilers with a new efficient gas boiler while maintaining the rest of the heating system.
Impact categories
The impact categories examined in the current study were Global warming potential (GWP), Eutrophication potential (EP), Fossil fuel depletion (FFD) and Human toxicity potential (HTP). Environmental impact assessment was performed according to CML 2001 methodology (Valente et al. 2011). The above considered impact categories (except EP) were chosen over more policy relevant indicators like particulate matter due to the representativeness of the impact of mining in infrastructure, so better meeting the goal of the study. However, EP was included as it is important to measure the impact on flora and fauna as NW Europe has large land water bodies (Bartolozzi et al. 2017).
Assumptions
The efficiency of the large boiler in the district heating system was assumed to an average of 90% over the lifetime, based on the GaBi database with the smaller boilers in the SDCC buildings assumed to be 85% over their lifetime including degradation (Thinkstep 2018). It was assumed that all raw materials were produced in the EU- 28 apart from cast iron. An assumption was made that the material is transported from Rotterdam port to Dublin Port to the system site. The only exception to this is the wood chips which would be sourced from Laois sawmills.
As there was limited information on waste heat being taken from a data centre, an assumption is made that the collector coil for the waste heat within the data centre exhaust is of a similar size to the heat exchanger. This assumption is based on both pieces of equipment performing the same function. The thermal store for the district heating systems was assumed to be 1.5 cm thick with insulation made up of high-density polyurethane (HDPU). Both systems were assumed to use gas boilers to provide 20% of the required heat. The transmission heat losses for the district heating systems were 3% for the Heat pump and 2% for the Biomass CHP, which was adopted from analysis carried out by Codema (2015). The decreased losses in the analysis were attributed to the superior system characteristics such as better insulation and capability to generate higher heat density.
For most of the equipment, dimensions and material type were sourced from the equipment manufacturers. Therefore, the assumed density of the materials was needed to calculate the weight of equipment. The density of steel, high density polyurethane (HDPU) and high-density polyethylene (HDPE) was assumed to be 7850 kg/m3, 950 kg/m3 and 100 kg/m3, respectively. Ammonia refrigerant in the heat pump was assumed to have no GWP impact (ASHRAE 2017).
Life cycle inventory
The life cycle inventory stage requires collection of input and output data for the studied system. To analyses the different DH systems, two types of data were used. Foreground data describes the data related to the inputs (electricity, fuel usage etc.,) for the heat pumps and gas boiler, along with the material used for manufacturing these equipment’s and building infrastructure. Background data refers to data that represents the generic materials, energy and transport involved in production processes and delivered to the foreground system as aggregated datasets. These datasets are generally taken from databases and literature. In this study, a 2016 average Irish grid electricity data was adopted from SEAI report (2018). All other fuel types were taken from GaBi’s professional database of Irish fuel emissions, as extraction and transport were included in these datasets. Data for the heat pump DH (WHP-DH) and gas boiler systems were both provided by the Codema (2015) report with the Biomass CHP district heating system (BCHP-DH) taken from Gartland (2014), Murphy et al (2014) and Murphy et al (2016). These data sources were used to define the three system’s equipment needs (Table 1) with fuel needed to produce one MWh of heat calculated from efficiency data. The piping systems employed in the study were adopted from Logstar catalogue (2018) and Codema (2015). Further, inventory related to pipes in terms of construction materials and dimensions considered in the model are shown in Table 2. When the system processes were defined, the background data for these heating systems were taken from Ecoinvent 3.7, GaBi professional database (Thinkstep 2018) and other published LCA studies (Table 1).
The data and sources for the WHP-DH are outlined in Table 1. The seasonal performance factor for the heat pump is assumed to be 3.6 with average 2016 Irish electricity used to power it. Heat losses in transmission are assumed to be 3% of total heat. The refrigerant used in the heat pump was ammonia.
The woodchips for the BCHP-DH system are harvested according to Scenario 3 proposed by Murphy et al. (2014). The energy of the woodchips at moisture content 35% is 11GJ/t which is adapted from Murphy et al. (2016). The wood at the sawmill has a moisture content of 20% and has an energy content of 12.65 GJ/t. Losses in wood during chipping is assumed to be 5%.
The wood is assumed to come from a large, forested area in Laois which is 80 km away from the CHP plant site. The data for the biomass boiler was taken from the gas boiler information for the district heating system as both were assumed to have heating capacity of 5 MW (Table 1). The building area which houses the biomass boiler was 100 m2 and the pipeline to the end users was 137 m with a heat exchanger substation. Allocation of burden to heat in the CHP was done according to the average Irish CHP efficiency, with heating allocated 62.8% of the burden and electricity burden excluded from this study.
For the gas boiler system (GB), the only infrastructure change was the gas boiler, with the quantity and size provided from Dublin’s Energy Agency (Codema). Efficiency of the boiler was assumed to be 85% with a connection to gas already established.