_____________________________________________________________________________________________________
BALANCING SUPPLY AND DEMAND IN
A DEDICATED BIOGAS GRID
_________________________________________________________________________________________________________________ Final Thesis, MSc Supply Chain Management
University of Groningen, Faculty of Economics and Business July 11th, 2018 DOEKE DIJKHUIZEN Student number: 2172690 d.t.dijkhuizen@student.rug.nl Supervisors – University Dr. M.J. Land Co-assessor – University Prof. Dr. Ir. J.C. Wortmann Supervisor – Field of study
B. van der Velde
i
“Balancing Supply and Demand In a Dedicated Biogas Grid”
ABSTRACT
The use of a dedicated infrastructure for transportation of biogas is gaining interest. Biogas producing farmers can be directly connected with pipelines to end-users. In these new networks balancing problems arise because gas demand is fluctuating while biogas production is a stable process. Research on the balancing issue that occur in these grids is scarce and this paper aims to fill that gap in current literature by answering the question: ‘How can buffering be used in dedicated biogas grid to balance supply and demand?’ A simulation study was conducted to see what the balancing performance of several buffer types is. Results of this study show that when the demand pattern has dips during the weekend a pressurized buffer is needed to capture supply during this periods of low demand. When the demand pattern is more continuous, buffer under the gas roofs of the digester and line-packing yield promising results.
ii TABLE OF CONTENTS
ABSTRACT ... i
1. INTRODUCTION ... 1
2. THEORETICAL BACKGROUND ... 3
2.1. Dedicated Biogas Grids ... 3
2.2. Biogas Utilization ... 4
2.3. Biogas Buffering ... 5
2.4. Conclusion ... 7
3. METHODOLOGY ... 8
3.1. Model, Variables and Parameters ... 8
3.2. Scenario’s ... 9 3.3. KPIs ... 10 4. CASE STUDY ... 12 4.1 Grid Dimensions ... 12 4.2 Supply Data ... 13 4.3 Demand Data ... 13 5. RESULTS ... 16 5.1 Current Situation ... 16
5.2 Influence of Increased Supply ... 19
5.3 Comparing Supply-Demand Ratios ... 22
5.4. Impact of Excluding Weekend Dip ... 24
6. DISCUSSION ... 28
7. CONCLUSIONS AND FUTURE RESEARCH ... 30
REFERENCES ... 31
1
1. INTRODUCTION
In the last two decades interest in renewable energy sources is growing. The European Commission for instance, issued a directive to achieve a 20% share of renewable energy by 2020 (European Union, 2009). More recently, Dutch politicians issued a request to raise taxes on natural gas with 75% to discourage gas consumption (Hofs, 2018). Biogas is a renewable substitute for natural gas, produced by digesting biomass (Holewa et al., 2013). Animal manure is a promising source of biomass because of the wide availability and high methane content (Koppejan and De Boer - Meulenman, 2005; Østergaard, 2012). Furthermore, digesting animal manure is also a way for farmers to mitigate the disposal challenges surrounding animal manure (Atandi and Rahman, 2012). New biogas infrastructures are emerging with decentralized production and dedicated biogas grids to connect producer and end-user as main components (Van Eekelen et al. 2012; Hengeveld et al., 2016).
In such dedicated grids, balancing issues arise because the supply of biogas is stable, while the demand fluctuates (Østergaard 2012; Trojanowska and Lipczyńska 2014). Besides seasonal variation, the gas demand also fluctuates on a weekly, daily and even hourly basis (Bekkering, Broekhuis and van Gemert, 2010). Bekkering et al. (2015) state that buffer against seasonal fluctuation is not a viable option because of the required buffer size. While Thompson et al. (2009) state that daily fluctuations should also be dealt with, and buffering can be of use. Hengeveld et al. (2014) state that research on the buffering function of the grid itself, so-called line-pack flexibility, should be conducted in a biogas setting. Bekkering, Broekhuis and van Gemert, (2010) suggest that when upgrading of biogas to natural gas quality is not necessary, seasonal as well as daily demand fluctuation are crucial to consider. So, buffering to handle seasonality has been treated in current literature. However, research on handling daily and weekly fluctuation is missing. To fill this gap in literature, this paper addresses the question: ‘How can buffering be used in dedicated biogas grid to balance supply and demand?’
2 from a local biogas grid in the Netherlands to analyze how demand fluctuations affect biogas utilization. In this biogas grid the aforementioned balancing problems occur and buffering can be an interesting instrument to tackle these. Furthermore, the performance of multiple buffers on balancing supply and demand is assessed with a simulation model. Gas flows in and out of the biogas system are analyzed with the use of diagrams to assess the balancing performance of multiple buffering scenarios. The balancing performance is determined with two KPIs: biogas that is flared off as percentage of total supply and natural gas consumption as percentage of total demand.
3
2. THEORETICAL BACKGROUND
In this section relevant literature for this study is discussed. First some papers that provide background information on these types of configurations are treated. Then, biogas end-uses are described and the implications these end-uses have on the balancing issues in biogas grids are discussed. Thereafter, literature on biogas buffering is described and multiple buffer types are presented. Last, a summary of the knowledge presented in this section is provided and the aforementioned research question is placed in light of these insights.
2.1. Dedicated Biogas Grids
Dedicated biogas grids solve a lot of the logistical problems that have been surrounding biogas (Bojesen, Birkin and Clarke, 2014). However, even with the use of a dedicated biogas grid, new problems arise because the anaerobic digestion process is producing biogas at a stable rate (Pöschl, Ward and Owende, 2010) while gas demand varies over time. Thus, a balancing problem of supply and demand emerges. This balancing problem needs to be solved, otherwise the produced biogas during periods of low demand would go to waste (Van Eekelen et al., 2012). Østergaard (2012) proposes installing a buffer to balance supply and demand.
4 issues did not occur because the biogas was upgraded. Van Eekelen et al. (2012) presented a report on the current state of biogas network and provided a detailed view on the costs associated with these grid. However, Van Eekelen et al. (2012) approach dedicated grids from an economical point of view and did not take the balancing issues into account, but merely described the technicalities of biogas infrastructures. Bekkering et al., (2013) focused on balancing seasonal demand in green gas grids on a regional scale, and concluded that flexible production by adding an extra digester during the winter was the most viable option. However, they did not include daily, weekly or monthly balancing issues.
Biogas grids are not well described in current literature, while many studies focus on the transportation of natural gas through pipelines. Natural gas is extracted from wells and transported through a network of high pressure pipelines over long distances (Vasconcelos et al., 2013). In order to deal with peak demand natural gas is also stored in large storage facilities (Thompson, Davison and Rasmussen, 2009; U.S. Energy Information Administration, 2015). Besides transportation, the natural gas grid is also used as storage facility, so-called line-packing (Keyaerts et al., 2010). While insights of natural gas system can be applicable to a biogas infrastructures, Hengeveld et al. (2016) stress the differences between biogas grids and natural gas grids. In biogas grids, the covered distance is smaller compared to natural gas grids. Furthermore, the gas volumes are smaller as well, while the number of extraction points is larger in biogas settings.
2.2. Biogas Utilization
5 economically viable for small scale grids (Hahn et al., 2014). Biogas for industrial process heat purposes is another market with high potential because of the scale and the possibility of using raw biogas (Coates, Moon, and Weaver, 2014). Furthermore, industrial gas demand tend be less temperature sensitive compared to domestic and commercial users (Vitullo et al., 2009). Instead of seasonality, industrial gas demand patterns are more affected by weekly of other periodic cycles (Vitullo, 2011). Using raw biogas decreases the total investment costs for the grid, because no upgrading installation is necessary which makes it an attractive alternative in combination with small scale grids. However, biogas infrastructures where raw biogas is used by the end-users has not yet been described in current literature. The disadvantage of this configuration is that the demand is constraining the output of the system, because gas can only be used when there is gas demand from the end-users that are connected. In times of low demand, production of biogas continues and needs to be stored to be utilized on a later moment. 2.3. Biogas Buffering
6 should be used to cover surplus supply. Donders, Van der Putten and Holstein (2010) also suggest that for longer periods of unbalance in the grid, buffering is the most suitable option. Hahn et al. (2014) argues that for larger gas buffers, external storage should be considered. A simple gas storage construction would be a large metal tank with valves to regulate pressure (Østergaard, 2012). Hahn et al. (2014) state that biogas can be stored under low, medium and high pressure. High pressure is not considered as a viable option, because of additional cooling and compressing requirements. Now, several of these biogas buffer types will be further explained.
2.3.1. Line-pack
Besides transport, the pipelines can also be used for storage (Hengeveld et al., 2016). This form of buffering is the so-called ‘line-pack flexibility’ and yields low costs. The line-pack buffer can be used to handle peak days (Coates, Moon and Weaver, 2014). The total capacity of the line-pack flexibility is the volume of gas that is the difference between the upper and lower pressure limit of the pipes (Keyaerts et al., 2010). The line pack is a function of the length of the grid, pipe diameter, pressure limits and flow rate. When the flow rate is higher, the line pack would decrease. The line-pack is particularly large in high pressure grids (Fevre, 2013). In large natural gas grids this line-pack is sufficient to balance fluctuations in gas demand (Keyaerts, 2012). The dimension of a biogas grids are small compared to the natural gas grid, and is expected to yield less line-pack flexibility.
2.3.2. Gas roof buffering
Another low cost form of buffering is storage under the gastight covering of the digester (Ertem and Acheampong, 2018). When the gas cannot be supplied, the gas will build up under the gas roof of the digester. The capacity of these gas roofs depend on the type of digester and the feedstock, but is on average 4-6 hours of biogas production for small digesters (Hahn et al., 2014) and up to 10 hours for large digesters (Ertem and Acheampong, 2018). This is quite some buffer capacity, because multiple digesters are connected in a micro grid. This makes the buffering of biogas within the digester a cheap and promising way to balance supply and demand in a micro grid.
2.3.3. Unpressurized external storage
7 Because these gas holding balloons are hold above the ground, additional safety requirements need to be in place (Deublein and Steinhauser, 2008).
2.3.4. Pressurized external storage
The pressure in a cylindrical storage tank can be increased in order to force the capacity up. This increase in capacity would yield higher utilization of the biogas. However, putting a lot of pressure on raw biogas has its downsides. Higher pressure means larger compressors. These larger compressors are more expensive than the smaller types. Furthermore, when biogas is compressed, condensed water can drop out of the biogas (Ray, 2016). Because the external biogas storage tanks are made of steel, this water could lead to corrosion and damage the storage tanks. To prevent corrosion from happening, additional cooling installations need to be installed (Van Eekelen et al., 2012). Obviously, the presence of these cooling installations increase investment costs as well. Besides higher investment costs, operational costs would increase significantly due to higher energy consumption of the larger installations.
Because of their large size and low costs, underground pressurized storage is most favorable for storing large quantities and/or for long periods of time (Apt et al., 2008). (Secomandi, 2010) states fuel is used when compressing for injection, not for withdrawals i.e. energy is only consumed when the buffer is filled, not when the buffer is deflated. 𝑁𝑚3
2.4. Conclusion
8
3. METHODOLOGY
In this section the methodology that is used for this research is described. First, the mathematical model that is used for simulation is presented. The simple simulation model is used to visualize gas flows in and out the system over time. Furthermore, KPIs are used to assess the performance of multiple buffer types. Second, the scenario’s used for simulation are described. In these scenario’s buffer types are combined to see which combination yields the most promising results.
3.1. Model, Variables and Parameters
3.1.1. Model
In order to graphically represent the flows of biogas in system over time, the model calculates how much biogas is flowing in and out of the system for each hour t. The expression that is used to determine a delta for each hour t is denoted by equation (1), where demand t is extracted from supply t. A negative ∆𝐼𝑡implies that the supply during t is not sufficient to satisfy all demand during t. When ∆𝐼𝑡 is negative biogas is extracted from the system. On the other hand, when ∆𝐼𝑡is positive the opposite is true and gas is pumped into the system. The total amount of biogas in the system during hour t is denoted by 𝐼𝑡. When 𝐼𝑡 is equal to zero, the system is empty, and no additional demand can be satisfied by the system. When during these moments ∆𝐼𝑡is negative, the biogas is complemented with natural gas, which is denoted by equation (2). When the system is completely filled e.g. 𝐼𝑡 is equal to the capacity of the system, while ∆𝐼𝑡 is positive, the biogas cannot be handled by the system and has to be flared for safety reasons. This is denoted by equation (3). Equation (4) denotes the total buffer capacity of the grid which is composed of line-pack flexibility, gas roof buffering and an external buffer.
∆𝐼𝑡 = 𝑆𝑡− 𝐷𝑡 (1)
∆𝐼𝑡+ 𝐼𝑡−1< 0 ⇒ 𝑁𝐶𝑡 = −∆𝐼𝑡− 𝐼𝑡−1 (2)
∆𝐼𝑡+ 𝐼𝑡−1> 𝐶 ⇒ 𝐹𝑡 = ∆𝐼𝑡+ 𝐼𝑡−1− 𝐶 (3)
𝐶 = 𝐿𝑃 + 𝐺𝑅 + 𝐵𝐶 (4)
3.1.2. Variables and Parameters
9 multiplying π with the squared radius and the grid length. Then this number needs to be multiplied with the difference between the upper and lower operating pressure limit. The gas roof buffer capacity is set at the average biogas supply of 4 hours.
TABLE 1
Description of Variables and Parameters
Variable / Parameter Description
𝑰𝒕 Biogas in system at t
𝑵𝑪𝒕 Natural gas consumption at t
𝑭𝒕 Flared biogas at t
𝑫𝒕 Demand at t
𝑺𝒕 Supply at t
C Total system capacity
LP Line-pack flexibility
GR Gas roof buffer capacity
BC Buffer capacity
3.2. Scenario’s
10
TABLE 2 Scenario Description
Scenario 1 2 3 4 5 6 7 8
Line-pack flexibility No Yes Yes Yes Yes Yes Yes Yes
Gas roof buffering No No Yes Yes Yes Yes Yes Yes External buffer (pressure) No No No Yes (1 bar) Yes (2 bar) Yes (3 bar) Yes (4 bar) Yes (8 bar) 3.3. KPIs
The performance of the buffer scenarios is assessed by the score on a set of KPIs. This section describes how these KPIs are calculated. Besides biogas utilization and natural gas consumption, CAPEX and OPEX are calculated to assess the impact of increasing the buffer capacity on the cost per 𝑁𝑚3of biogas. For this costs calculations, generally accepted numbers from leading reports on biogas infrastructures are used to produce a well-grounded estimation.
3.4.1. Biogas Utilization
The first KPI is the utilization rate of the biogas. This is the percentage of total biogas supply that is used by the consumers, shown in equation (5). This is determined by the amount of gas that is flared, which is expressed by equation (3) in the section 3.1.
𝑈𝑡 =∑𝑆∑𝑆𝑡−∑ 𝐹𝑡
𝑡
(5)
3.4.2. Natural Gas Consumption
When the system is not able to satisfy demand, the end-users switch to consumption of natural gas. The balancing performance of the grid is also assessed with the natural gas consumption as percentage of total demand.
3.4.3. CAPEX
11 atmospheric pressure with €222,-. This calculation yields a estimation of the capex that is considered to be sufficiently accurate. The costs for a compression installation for low to medium pressure are used from a report from E Kwadraat (2011) and are equal to €136.000, - including all needed installations. The depreciation period is set at 12 years, which corresponds with the period used by subsidy schemes of the Dutch government (Hengeveld et al., 2014).
3.4.4. OPEX
The operational costs of a biogas buffer are determined by the operational costs of the compressor and the cooling installation. These costs are made up from maintenance costs and energy use. When the pressure is increased, the compressor uses more energy, the same applies to the cooling installation. To calculate the energy use of the compressor, the equation from Van Eekelen et al. (2102) in the Appendix is used. The maintenance costs are 3% of CAPEX per year (Bärnthaler et al., 2008).
3.4. Comparison of Supply-Demand Ratios
12
4. CASE STUDY
The case that is analyzed with the presented model is a dedicated biogas grid in the Netherland. This novel grid connects seven biogas producing dairy farms with two industrial end-users. The biogas is injected in the grid at a steady rate. This rate differ a little during the year because some farmers use a grazing regime during the summer months, which decreases the amount of manure that is available.
4.1 Grid Dimensions
To calculate some of the parameters data of the grid dimension is needed. These dimensions contain the total length of the grid (11,000 meter), the diameter of the pipes (150 mm), the upper (300 mbar) and lower (125 mbar) pressure limit. With these dimensions the line-pack flexibility of the grid can be calculated. The buffer capacity of each scenario is presented in Table 3.
TABLE 3 Scenario Description
Scenario 1 2 3 4 5 6 7 8
Line-pack flexibility 0 55 55 55 55 55 55 55 Gas roof buffering 0 419 419 419 419 419 419 419 External buffer 0 0 163 652 1303 1955 2607 5213 Total (𝑁𝑚3) 0 474 637 1126 1777 2429 3081 5687
4.1.2. CAPEX
13
4.1.2. OPEX
The operational costs are composed of maintenance costs and energy consumption. The energy consumption of the compressor can be calculated with the expression in the Appendix. This results in an energy consumption of 0.21 kWh/𝑁𝑚3 for compression to 1 bar, to 0.35 kWh/𝑁𝑚3for compression to 8 bar in the buffer. The price for 1 kWh is set at €0,12. The maintenance costs are 3% of the total CAPEX per year, which is equal to €10.740,- per year. Finally, the total CAPEX and OPEX is divided by the biogas that is supplied, so excluding the flared volumes, to calculate a costs price per 𝑁𝑚3 of biogas.
4.2 Supply Data
The supply data is used as input for the model as presented in section 3. To represent the minor seasonal variation in biogas supply during the year the supply follows the distribution presented in Table 4. The total biogas supply is 1,500,000 𝑁𝑚3 per year. However, one 𝑁𝑚3 of biogas cannot satisfy one 𝑁𝑚3 of natural gas supply. Because the biogas produced has a methane
content of 58% while natural gas contains 89% methane, the biogas supply has to be multiplied with a 58/89 ratio. I.e. one 𝑁𝑚3 of biogas is equal to roughly 1.5 𝑁𝑚3 of natural gas. Which
comes down to a biogas production of 977,528 𝑁𝑚3 natural gas equivalent.
TABLE 4
Yearly Supply Distribution
Month Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec Percentage
of average 105% 105% 105% 100% 95% 95% 95% 95% 95% 100% 105% 105%
4.3 Demand Data
14
FIGURE 1
Yearly Supply and Demand Pattern
Figure 2 provides a more detailed view on this weekly pattern. The graph shows actual hourly gas demand during the first full week of 2014. What the graph clearly shows is the drop in demand during the weekend. During the weekdays the demand fluctuates heavily from hour-to-hour. In a closed system as this, these hourly fluctuation can already cause problems with biogas supply. Furthermore, the drop in demand during the weekends causes a huge surplus of supply. When this surplus of supply is not handled properly, this biogas would go to waste. This production surplus during the weekend is crucial to consider when determining the buffer capacity. This is in line with findings of Vitullo et al. (2009), they disaggregated gas demand and found that industrial users tend to have less gas demand during the weekend.
During the week demand is higher than supply for most of the time, except during the weekends. The total gas demand over 2014 was 1,238,622 𝑁𝑚3, and over 2015 it was 1,241,243 𝑁𝑚3.
15
FIGURE 2
16
5. RESULTS
In this section the results of the scenario analysis are presented. First the results of the scenarios in the current state are presented. The gas flows in the system are presented in graphs and buffer performance in terms of flared volume and natural gas consumption is shown. Second, the same scenarios are tested for the hypothetical situation where total supply equals total demand. The same procedure is used to present these results. Third, the results of the effect supply-demand ratios have on the performance of the buffer are presented.
5.1 Current Situation
The graph in Figure 3 depicts the simulation results of scenario 7 as described in the previous section. In this scenario the buffer capacity of the gas roofs, line-pack and a pressurized buffer under 4 bar are combined. The capacities of these buffers are depicted by the horizontal lines in the graph. The curve shows the amount of biogas that is in the system during time t in the first two weeks of February 2014. During this period, the gas demand is at its highest point in the year. From the graph a weekly pattern can be identified. The system is constantly filled during the weekends because demand is low and a surplus of supply is captured by the buffer. When the weekend is over, the end-users start producing again and the system is completely deflated by a surplus of demand. When the system is completely emptied, the end-users switch to natural gas to satisfy their demand. Thus, at those moments natural gas supplements biogas which is shown as a negative amount of gas in the system.
17 The results for all scenarios are provided in Table 5. It is clear that without any buffer capacity the biogas utilization is surprisingly high, 81.35%. This is because demand is higher than supply for all weekdays except the weekends. Adding the buffer capacity of the line-pack flexibility yields no significant gains in biogas utilization. Because of the dimensions of the small grid that is analyzed, the capacity of the line-pack flexibility is not sufficient to deal with the weekly fluctuations in demand. Adding the buffer capacity of the gas roofs has more impact on biogas utilization. Without any additional costs, the utilization rate rises almost 3%. However, to realize significant results, external buffer capacity is necessary. The costs per 𝑁𝑚3 of the pressurized buffers are decreasing when the pressure increases, only the OPEX per 𝑁𝑚3 increase a little bit when pressure is over 4 bar. Increasing the buffer capacity by increasing the pressure is effective in terms of biogas utilization. The 8 bar buffer almost yields a utilization rate of 100%. However, the effect of adding buffer capacity diminishes when utilization gets close to 100%. At first, adding 1 bar of pressure to the storage vessel (from 1 to 2 bar) leads to an increase of utilization of more than 3%, while increasing pressure from 4 to 8 bar almost 2%.
TABLE 5
Simulation Results Scenario’s
Scenario 1 2 3 4 5 6 7 8 Biogas Utilization 81.35% 82.60% 85.33% 90.17% 93.45% 96.17% 98.02% 99.82% Flared biogas 364501 18.65% 340136 17.40% 292599 14.67% 180323 9.83% 120141 6.55% 70313 3.83% 36408 1.98% 3266 0.18% Natural Gas consumption 964923 38.91% 891889 35.96% 818780 33.28% 825723 33.29% 765542 30.87% 715714 28.86% 681808 27.49% 648667 26.16% CAPEX (€/𝑁𝑚3) 0 0 0 0.2284 0.1193 0.0844 0.0687 0.0565 OPEX (€/𝑁𝑚3) 0 0 0 0.0095 0.0067 0.0060 0.0057 0.0060
18 Figure 4 shows, the curves for buffer above 3 bar are all the same. Which means that increasing the pressure beyond 3 bars won’t have any additional effect on biogas utilization, at least during winter.
FIGURE 4
Comparison of Buffer Capacities for Current Situation
In the second graph in Figure 4 the same curves are plotted for the first two weeks of August. During summer the demand for gas is lower, which increases the production surplus during the weekends. Furthermore, there is less surplus of demand during the weekdays which deflates the buffer slower during the week. During summer months increasing the buffer has more effect than during winter. Additionally, the first graph in figure 4 shows no curves for pressures above 3 bar. This is because the graphs for 4 and 8 bar are equal to the graph of 3 bar. This implies
First two weeks of February
19 that during winter increasing the pressure beyond 3 bar yielded no positive anymore. During summer the 8 bar buffer gains is of more use. Still the buffer is completely emptied during the week. However, the buffer is filled up further than during winter. The 8 bar buffer is never completely filled during these weeks, and utilizes significant volumes of biogas. The buffers under the lower pressure are completely full during weekends (shown by the flat tops of the curves), which means biogas has to be flared off to prevent overpressure.
Finally, the goal seek solver is used in Excel to find the minimum value of the total buffer capacity that yields a biogas utilization of 100%. The solver produces a total buffer capacity of 10,379 𝑁𝑚3, which implies the pressure in the storage vessel must be raised to 15 bar.
5.2 Influence of Increased Supply
Besides the variation of demand the total amount of supply is determining the required buffer size. In the actual situation average demand was higher than average supply. Because demand is so high compared to supply, a completely filled buffer is fully emptied within one day. To see what the impact of more supply for the same level of demand is, the supply is raised until it is equal to demand on average.
FIGURE 5
Simulation Results Scenario 7
20 line-pack and pressurized 4 bar buffer are combined. The graph shows the amount of biogas that is in the system for the first four weeks of January 2014. Again, a dominant weekly pattern can be identified. Because the demand remained the same, the weekend dips are still present. The higher level of supply increases the need for buffering because more biogas needs to be stored during the weekend. The buffer is completely filled during the weekend and biogas has to be flared to prevent overpressure. The 4 bar buffer is not able to capture all supply. After being completely filled, the system is completely deflated again in a short period of time. After being completely deflated, the system is not able to satisfy demand anymore, leading to consumption of natural gas on weekdays. The dips in demand during the weekends still dominate the system. Supply is equal to demand on average, but the skew distribution of demand in the weekends leads to a large surplus of supply during the weekends, while during the weekdays there is a shortage.
TABLE 6
Simulation Results when Supply = Demand
Scenario 1 2 3 4 5 6 7 8 Biogas Utilization 75.47% 80.25% 81.95% 83.02% 85.70% 88.26% 90.64% 94.86% Flared biogas 607812 24.53% 489572 19.75% 472004 19.05% 420781 16.98% 354386 14.30% 291019 11.87% 231947 9.36% 127429 5.14% Natural Gas consumption 609546 24.58% 491807 19.83% 474403 19.13% 423374 17.07% 357061 14.40% 294346 11.87% 235926 9.51% 134015 5.40% CAPEX (€/𝑁𝑚3) 0 0 0 0.212 0.106 0.072 0.055 0.037 OPEX (€/𝑁𝑚3) 0 0 0 0.0059 0.0047 0.0045 0.0045 0.0049
Table 6 shows the results of the simulation for each scenario with average supply equal to average demand. The utilization levels are lower compared to the results in Table 5. However, the natural gas consumption is significantly lower. Furthermore, the CAPEX per 𝑁𝑚3 as well as the OPEX per 𝑁𝑚3 decreases compared to the initial situation. This is because the total costs are spread over a larger volume of biogas.
21 Figure 4. Because the higher supply level, more biogas needs to be buffered during the weekends. In the initial situation a 3 bar buffer was sufficient for the amount that has to be buffered. When supply is raised, the amount of biogas that needs to be buffered during the weekend increases, which implies that the required buffer increases as well. The second graphs presents the results for the first two weeks of August 2014. During winter the buffer is constantly completely deflated. During summer a different pattern occurs because the system is not always completely deflated anymore. This is best shown by the curve for the 8 bar buffer. During this summer period demand is not high enough to empty the buffer. All buffer types are completely filled during the weekend, yielding additional flared volumes of biogas which decrease with increasing pressure. On the other hand, the amount of natural gas that is consumed is decreased. In the scenario of an 8 bar buffer, there is even no natural gas consumption anymore during this time period.
FIGURE 6
Comparison of Buffer Capacities when Supply=Demand
First two weeks of February
22 Again, the goal seek solver in Excel is used to determine the minimum buffer capacity that yields a 100% biogas utilization. When average supply is equal to average demand, the solver produces a required buffer capacity of 74,612 𝑁𝑚3, which implies the pressure in the storage vessel should be over 110 bar. This pressure level is considered to be unfeasible for biogas configurations.
5.3 Comparing Supply-Demand Ratios
To further investigate the effect of how supply compares to demand, multiple supply-demand ratios are tested. The effect of these ratios is assessed on flared biogas as percentage of total supply and natural gas consumption as percentage of total demand. Average supply is entered in the model as ratio from 0.1 to 1.0 of average demand.
Figure 7 shows the results of this analysis. The graph begins at a supply-demand ratio of 0.4 because before this ratio no biogas is flared. Thus, demand is constantly higher than supply when the supply-demand ratio is smaller than 0.4. As the ratio increases, the percentage of flared biogas increases as well. This means that when the supply-demand ratio increases and a certain percentage of flared biogas needs to be maintained, the buffer capacity needs to increase as well. Furthermore, the graph shows that increasing the buffer capacity has a declining effect on the percentage of flared biogas. The difference in performance between a 7 and 8 bar buffer is significantly smaller than the difference between 1 and 2 bar.
FIGURE 7
Supply-Demand Ratio Comparison, Flared Biogas
23 curves in Figure 7. Increasing the supply-demand ratio negatively impacts the consumption of natural gas e.g. more biogas supply leads to less natural gas consumption. Again the effect of increasing the buffer capacity is declining. Raising the pressure beyond 8 bars would yield a lower decrease in natural gas consumption than the increase from 1 to 2 bar. Furthermore, a 0% natural gas consumption is not reached. This has to do with the seasonality of gas demand. The final 5% of flared biogas has to be stored during summer, to satisfy the final 5% of natural gas consumption during the winter. The buffer would need to be extremely large to overcome this seasonality.
FIGURE 8
Supply-Demand Ratio Comparison, Natural Gas Consumption
24
FIGURE 9
Supply-Demand Ratio Comparison, Combined
Because demand during the weekends is so low, the buffer need is dominated by these periods of low demand. The buffer is not sized to deal with hourly fluctuations in demand, but is mainly needed to deal with the supply surplus during the weekends. This reoccurring weekly pattern heavily impacts the buffer need in the system. To determine the required buffer size for such a demand pattern, practitioners should multiply the highest hourly production with 48 to be able to handle the weekend dip. However, for some systems this demand pattern is not applicable. Industrial consumers might produce during all days of the week. To assess the impact of continuous production by end-user during the week, the weekends are removed from the dataset.
5.4. Impact of Excluding Weekend Dip
The weekly pattern in demand is dominating the buffer need for this system. Because demand is so low during the weekend, the buffer need to be sized to capture the supply during the weekend, rather than to deal with hourly fluctuation. Furthermore, industrial end-users may produce seven days a week e.g. have a continuous demand pattern.
25
FIGURE 10
Weekly Demand Distribution without Weekend Dip
FIGURE 11
Simulation Results, Excluding Weekend Dips for Initial Supply
26 any buffer capacity the biogas is utilized for 94.31%. Adding the capacity of the digester roof and the line-pack flexibility already yields a utilization of 98.22%. The supply-demand ratio is roughly 0.6 in this situation, which impacts the required buffer size.
TABLE 7
Simulation Results, Excluding Weekend Dips for Initial Supply
To further analyze the impact of a more continuous demand pattern the supply is raised till the supply-demand ratio reaches 1 e.g. average supply is equal to average demand. When average supply is equal to average demand, the required buffer size increases. Figure 12 shows the biogas volumes in the system, again during the first weeks of February 2016, for scenario 3. Because of the higher level of supply, the buffer capacity of the digester roofs is increased as well (the capacity is set as 4 times the average hourly supply). However, the buffer capacity of the line-pack and the gas roofs is not sufficient to capture all supply anymore. During summer the buffer is completely filled and large volumes of biogas are flared. During winter an opposite pattern is true, the buffer is completely deflated and natural gas in consumed.
27
FIGURE 12
Simulation Results, Excluding Weekend Dips for Supply=Demand
The results of these simulations are presented in Table 8. The combination of line-pack and gas roof buffer is able to handle most of supply, namely 91.77%. Increasing the buffer capacity by an external buffer yields positive results, up till a certain point. From a point of 94% onwards, the seasonality becomes more of an obstacle. To reach 100% biogas utilization for this setting the Excel solver produces a required buffer capacity of more than 72,000 𝑁𝑚3, which implies the pressure in the external buffer needs to be over 100 bar.
TABLE 8
Simulation Results, Excluding Weekend Dips for Supply=Demand
28
6. DISCUSSION
This research aimed to visualize gas flows in and out of a dedicated biogas grid and provide insights in how these flows can be balanced by buffering. The results presented in the previous section show that the required buffer capacity is heavily depend on the demand pattern. In the case that was analyzed the weekend dips in demand are dominant in determining the required buffer capacity. The methods used in this paper provided insights in these weekend dips. Which is in line with findings of Vitullo (2011) who states that industrial gas demand decreases during weekends. The weekly pattern is dominating the buffer need and sizing the buffer to these weekend dips yields promising results. The daily and hourly demand variations are offset by the weekend dips in determining the buffer size. Therefore, it can be concluded that when the gas demand pattern in a dedicated biogas grid follows this weekly distribution with weekend dips, the buffer capacity of the grid should be set at 48 times the hourly biogas production e.g. the total production during the weekend. 48 might not be fully accurate to describe a full weekend, because workdays do not start and end at midnight. However, because there is still some demand during the weekends, this offsets these additional hours. In the initial situation this would imply that the total buffer capacity would need to be 5355 𝑁𝑚3 which is a little
under the total capacity of scenario 8 and would yield a utilization of around 99%. The actual supply level than determines the absolute capacity, which subsequently can be translated to a specific buffer type by determining the capacity for each buffer type.
29 reach a 100% utilization level, the gas that is flared during summer should be used during the winter months. To realize this, the pressure would need to be increased over 100 bar. This is in line with findings of Bekkering et al. (2013), who stated that buffering against seasonality is not a viable option.
In later research Bekkering et al. (2015) argued that it would be interesting to see if line-pack flexibility could be an feasible form of buffering in dedicated biogas grids, because it can be operated under low costs. Also Hengeveld et al. (2014) pointed out that research on balancing biogas grids with the buffer capacity of the grid would be interesting. However, this research shows that the capacity of line-pack flexibility is not sufficient to balance these grids. The dimensions of the grid, in combination with the pressure under which it’s operated produce a very limited buffer capacity. Van Eekelen et al. (2012) analyzed three cases with lengths of 3.6, 8, and 29 km. Calculations of O’Shea et al. (2017) yielded gird lengths between 14 and 32 km. Thus, the grid length of 11 km that was used in the analyzed case can be seen as representative for biogas grids in general. Therefore, line-pack flexibility would not be an interesting option for balancing biogas grids with decentralized production and small diameter pipes.
30
7. CONCLUSIONS AND FUTURE RESEARCH
In this paper several buffer options were translated into multiple scenarios to set the parameters for a simple simulation model. The results show that line-pack flexibility cannot be considered as a useful buffer type in dedicated biogas micro-grids. Simply because of the size of these grid, the capacity of the line-pack flexibility is too limited to play a relevant role in balancing supply and demand. However, line-pack flexibility can be operated without additional costs and should therefore be used. Buffering under the gas roofs of the digesters is a promising buffer type. Because it can be operated under low costs, and the investments are simply not needed, this buffer type is a good way to increase biogas utilization while keeping the costs low.
A medium pressure buffer can balance supply and demand properly. Raising pressure in a buffer, raises the storage capacity, more capacity leads to more utilization. However, the pressure in a buffer cannot exceed 10 bar, because of the technical difficulties that would occur. When industrial end-users in a dedicated biogas grid do not produce during the weekends e.g. have no gas demand during the weekend, the required buffer capacity should at least be 48 times the hourly biogas production. This can be used a rule of thumb for practitioners to determine the required buffer capacity. The outcome of multiplying the hourly production with 48 can be translated into a required buffer package.
When weekend dips are present buffering under the roofs of the digester and by means of line-pack flexibility yields insufficient buffer capacity. When demand is distributed continuously over the week, without weekend dips, the buffer need is heavily reduced and buffering under the digester roofs can already balance out most of the variation.
31
REFERENCES
Apt, J. et al. (2008) ‘An Engineering-Economic Analysis of Syngas Storage’, U.S. Department of Energy’s National Energy Technology Laboratory, p. 143. Available at: http://repository.cmu.edu/cgi/viewcontent.cgi?article=2129&context=tepper (Accessed: 11 June 2018).
Atandi, E. and Rahman, S. (2012) ‘Prospect of anaerobic co-digestion of dairy manure: a review’, Environmental Technology Reviews. Taylor & Francis, 1(1), pp. 127–135. doi: 10.1080/09593330.2012.698654.
Bärnthaler, J. et al. (2008) ‘Biogas-Mikronetze eine Initiative des Bundesministeriums für Verkehr, Innovation und Technologie (BMVIT) Gasversorgung mittels lokaler
Biogas-Mikronetze’. Available at:
http://www.biogas-netzeinspeisung.at/downloads/20080930_endbericht_mikronetze.pdf (Accessed: 17 May 2018).
Bekkering, J. et al. (2013) ‘Balancing gas supply and demand with a sustainable gas supply chain - A study based on field data’, Applied Energy. Elsevier, 111, pp. 842–852. doi: 10.1016/j.apenergy.2013.05.073.
Bekkering, J. et al. (2015) ‘Designing a green gas supply to meet regional seasonal demand – An operations research case study’, Applied Energy. Elsevier, 143, pp. 348–358. doi: 10.1016/J.APENERGY.2015.01.034.
Bekkering, J., Broekhuis, A. A. and van Gemert, W. J. T. (2010) ‘Optimisation of a green gas supply chain – A review’, Bioresource Technology. Elsevier, 101(2), pp. 450–456. doi: 10.1016/J.BIORTECH.2009.08.106.
Bojesen, M., Birkin, M. and Clarke, G. (2014) ‘Spatial competition for biogas production using insights from retail location models’, Energy. Pergamon, 68, pp. 617–628. doi: 10.1016/J.ENERGY.2013.12.039.
Börjesson, M. and Ahlgren, E. O. (2012) ‘Cost-effective biogas utilisation - A modelling assessment of gas infrastructural options in a regional energy system’, Energy. Pergamon, 48(1), pp. 212–226. doi: 10.1016/j.energy.2012.06.058.
Budzianowski, W. M. and Brodacka, M. (2017) ‘Biomethane storage: Evaluation of technologies, end uses, business models, and sustainability’, Energy Conversion and Management. Pergamon, 141, pp. 254–273. doi: 10.1016/j.enconman.2016.08.071.
Coates, S., Moon, B. and Weaver, D. (2014) ‘Long term gas supply and demand scenarios’, (September), p. 157. Available at: http://gasindustry.co.nz/dmsdocument/4771.
Deublein, D. and Steinhauser, A. (2008) Biogas from Waste and Renewable Resources - An Introduction. Wiley-VCH.
Donders, R., Van der Putten, E. and Holstein, J. (2010) Groen Gas Invoeding in het Gasnet. Available at: http://4solarenergy2.nl/downloads/02862010-11-08-groen-gas-invoeding-in-het-gasn.pdf (Accessed: 10 March 2018).
32 Van Eekelen, R. N. et al. (2012) ‘New Networks for Biogas - Current Practices, Risks, Cost Aspects’, p. 89. Available at: www.kiwagastechnology.nl (Accessed: 1 March 2018).
Ertem, F. C. and Acheampong, M. (2018) ‘Impacts of Demand-Driven Energy Production Concept on the Heat Utilization Efficiency at Biogas Plants: Heat Waste and Flexible Heat Production’, Process Integration and Optimization for Sustainability. Springer Singapore, 2(1), pp. 1–16. doi: 10.1007/s41660-017-0024-z.
European Union (2009) Directive 2009/28/EC of the European Parliament and of the Council. Available at: http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:32009L0028 (Accessed: 5 March 2018).
Fevre, C. Le (2013) Gas storage in Great Britain. Available at: http://infrastructure.planningportal.gov.uk/wp-content/ipc/uploads/projects/EN030001/3. Post Decision Information/Decision/130409_EN030001_Chris_Le_Fevre_Report.pdf (Accessed: 22 March 2018).
Hahn, H., Ganagin, W., et al. (2014) ‘Cost analysis of concepts for a demand oriented biogas supply for flexible power generation’, Bioresource Technology. Elsevier, 170, pp. 211–220. doi: 10.1016/j.biortech.2014.07.085.
Hahn, H., Krautkremer, B., et al. (2014) ‘Review of concepts for a demand-driven biogas supply for flexible power generation’, Renewable and Sustainable Energy Reviews. Pergamon, 29, pp. 383–393. doi: 10.1016/j.rser.2013.08.085.
Hakawati, R. et al. (2017) ‘What is the most energy efficient route for biogas utilization: Heat, electricity or transport?’, Applied Energy. Elsevier, 206, pp. 1076–1087. doi: 10.1016/J.APENERGY.2017.08.068.
Hengeveld, E. J. et al. (2014) ‘When does decentralized production of biogas and centralized upgrading and injection into the natural gas grid make sense?’, Biomass and Bioenergy. Pergamon, 67, pp. 363–371. doi: 10.1016/J.BIOMBIOE.2014.05.017.
Hengeveld, E. J. et al. (2016) ‘Biogas infrastructures from farm to regional scale, prospects of biogas transport grids’, Biomass and Bioenergy. Pergamon, 86, pp. 43–52. doi: 10.1016/j.biombioe.2016.01.005.
Hofs, Y. (2018) ‘Wie niet van het aardgas af wil, zal dit voelen in de portemonnee: klimaattafel van Samsom zorgt voor forse stijging van aardgasprijzen | De Volkskrant’, de Volkskrant, 21 June. Available at: https://www.volkskrant.nl/nieuws-achtergrond/wie-niet-van-het-aardgas- af-wil-zal-dit-voelen-in-de-portemonnee-klimaattafel-van-samsom-zorgt-voor-forse-stijging-van-aardgasprijzen-~ba2f343f/ (Accessed: 21 June 2018).
Holewa, J., Król, A. and Kukulska-Zajac, E. (2013) ‘Biogas as an alternative to natural gas?’, Chemik, 67(11), pp. 1073–1078. doi: 10.3103/S0967091209060096.
Keyaerts, N. et al. (2010) ‘Gas Balancing Rules Must Take into account the Trade-off between Offering Pipeline Transport and Pipeline Flexibility in Liberalized Gas Markets’, Vasa, (September). Available at: https://dspace.mit.edu/bitstream/handle/1721.1/59469/2010-014.pdf;sequence=1 (Accessed: 22 May 2018).
33
Available at:
https://lirias.kuleuven.be/bitstream/123456789/352326/1/Keyaerts_PhD_public_26-09-2012_final.pdf (Accessed: 24 March 2018).
Klosterhalfen, S. T., Holzhauer, F. and Fleischmann, M. (2018) ‘Control of a continuous production inventory system with production quantity restrictions’, European Journal of Operational Research. North-Holland, 268(2), pp. 569–581. doi: 10.1016/j.ejor.2018.02.001. Koppejan, J. and De Boer - Meulenman, P. D. M. (2005) De Verwachte Beschikbaarheid Van Biomassa in 2010 (The Expected Availability of Biomass in 2010).
Marsh, M. et al. (2005) ‘Biomethane from Dairy Waste A Sourcebook for the Production and Use of’, Prepared for Western United Dairymen. Funded part through USDA Rural Dev, (July),
pp. 1–282. Available at:
http://suscon.org/cowpower/biomethaneSourcebook/biomethanesourcebook.php (Accessed: 26 June 2018).
O’Shea, R., Wall, D. M. and Murphy, J. D. (2017) ‘An energy and greenhouse gas comparison of centralised biogas production with road haulage of pig slurry, and decentralised biogas production with biogas transportation in a low-pressure pipe network’, Applied Energy. Elsevier, 208, pp. 108–122. doi: 10.1016/J.APENERGY.2017.10.045.
Østergaard, P. A. (2012) ‘Comparing electricity, heat and biogas storages’ impacts on renewable energy integration’, Energy. Pergamon, 37(1), pp. 255–262. doi: 10.1016/j.energy.2011.11.039.
Patrizio, P. et al. (2015) ‘Biomethane as transport fuel - A comparison with other biogas utilization pathways in northern Italy’, Applied Energy. Elsevier, 157, pp. 25–34. doi: 10.1016/j.apenergy.2015.07.074.
Pierie, F. et al. (2016) ‘A new approach for measuring the environmental sustainability of renewable energy production systems: Focused on the modelling of green gas production pathways’, Applied Energy. Elsevier, 162, pp. 131–138. doi: 10.1016/J.APENERGY.2015.10.037.
Pöschl, M., Ward, S. and Owende, P. (2010) ‘Evaluation of energy efficiency of various biogas production and utilization pathways’, Applied Energy. Elsevier, 87(11), pp. 3305–3321. doi: 10.1016/J.APENERGY.2010.05.011.
Ray, nayak himanshu sekhar (2016) ‘Biogas Compression and Storage System for Cooking Applications in Rural Households’, International Journal of Renewable Energy Research (IJRER), 6(2), pp. 593–598.
Rupf, G. V. et al. (2017) ‘Development of an optimal biogas system design model for Sub-Saharan Africa with case studies from Kenya and Cameroon’, Renewable Energy. Pergamon, 109, pp. 586–601. doi: 10.1016/J.RENENE.2017.03.048.
Secomandi, N. (2010) ‘Optimal Commodity Trading with a Capacitated Storage Asset’, Management Science, 56(3), pp. 449–467. doi: 10.1287/mnsc.1090.1049.
34 U.S. Energy Information Administration (2015) The Basics of Underground Natural Gas Storage - U.S. Energy Information Administration, November 16 . Available at: https://www.eia.gov/naturalgas/storage/basics/ (Accessed: 22 March 2018).
Ullah Khan, I. et al. (2017) ‘Biogas as a renewable energy fuel – A review of biogas upgrading, utilisation and storage’, Energy Conversion and Management. Pergamon, 150, pp. 277–294. doi: 10.1016/j.enconman.2017.08.035.
Vasconcelos, C. D. et al. (2013) ‘Network flows modeling applied to the natural gas pipeline in Brazil’, Journal of Natural Gas Science and Engineering. Elsevier, 14, pp. 211–224. doi: 10.1016/j.jngse.2013.07.001.
Vitullo, S. (2011) ‘Disaggregating Time Series Data for Energy Consumption by Aggregate
and Individual Customer’, p. 151. Available at:
https://epublications.marquette.edu/dissertations_mu/169/ (Accessed: 8 July 2018).
35
