assessment of geothermal resources in xi'an, china - Orkustofnun
assessment of geothermal resources in xi'an, china - Orkustofnun assessment of geothermal resources in xi'an, china - Orkustofnun
Yin Lihe 424Report 20Case II: Four reinjection wells,instead of two as in case I, with aconstant 10 l/s injection rate lasting3 months during each year’swintertime. The injection rate is15% of the production rate duringwintertime of the year 2000. Figure11 shows the water level recoverycalculated by LUMPFIT.The results indicate that, with 2reinjection wells, the water level inthe system will rise about 7 mmaximum and 2 m minimum with 5m in average, long-term. With 4reinjection wells, the average waterlevel recovery will be 8 m, and theminimum and maximum water levelrecovery will be 5 and 14 m,respectively. There is no doubt thatthe production potential and lifetimeof the reservoir will improveaccordingly.Water level recovery (m)16128402 reinjection wells4 reinjection wells2000 2002 2004 2006 2008 2010FIGURE 11: Predicted water level recoverywith 2 and 4 reinjection wells4.3 Thermal breakthrough calculationsAll methods, regardless of how promising they may be, have their positive and negative sides. Besidescost consideration, thermal breakthrough is the most serious problem facing injection. Even thoughthermal breakthrough and cooling of reservoir fluid have not been major problems in any geothermalfields, it is necessary to predict thermal breakthrough time for different injection-production well spacing,i.e. the time from initial injection until a significant cooling is observed in a production well (Stefánsson,1997).Consider one reinjection well without a production well nearby. The injected water diffuses radially awayfrom the injection well through the porous rock matrix. If an intergranular flow is assumed, then the rockand the fluid have the same temperature at any point. The differential equation that describes this heattransport is∂Tβw+ q ⋅∇T= 0(7)∂tρβwhere T = Temperature (°C);$ w = Heat capacity of water (J/kg/°C); = Wet rock heat capacity (J/°Cm 3 );q = (q x , q y , q z ) the mass flux vector (kg/m 2 /s);LT = The temperature gradient vector.An infinite horizontal reservoir of constant thickness H is assumed. Injection of Q kg/s of cold water isassumed to start at time t = 0. The cold front moves away from the injection well, and the radial distancefrom the well to the temperature front is given as follows:
Report 20425Yin Lihe⎡ β ⎤r = wQtT ⎢ ⎥⎣πHρβ ⎦12(8)This formulation indicates that the radial distance is closely related to injection rate and time passed. Thenchanging the equation to an equation giving cold front break-through time as a function of the distancebetween reinjection and production wells, results in:2r0πHρβt =Qβ w(9)Three different injection rates, 10, 20and 30 kg/s of 10°C cold water, wereapplied to predict the breakthroughtime. The results calculated by theabove formulation are presented inFigure 12. It should be noted that thethickness of the reservoir used in theabove equation is 175 m, one fourth ofthe actual thickness of the secondformation, since most wells are locatedin that part of the reservoir.Based on the calculation, the distancebetween a production well and areinjection well should be longer than1000 m in order to prevent potentialthermal breakthrough for a very longtime. Because the reservoir iscomposed of alternate layers ofsandstone and mudstone, it is quitepossible that the injected water mighttravel along a thin sandstone layer withabnormal permeability, reducingcooling time dramatically.Break-through time(year)120010008006004002000Injection rate=10kg/gInjection rate=20kg/sInjection rate=30kg/s0 200 400 600 800 1000Distance (m)FIGURE 12: Estimated cold front breakthrough time as afunction of distance between reinjection and production well5. RECOMMENDED DESIGN OF A TRACER TEST5.1 Tracer testThe thermal breakthrough time estimated earlier depends on the channel geometry of the channelsconnecting the production and the reinjection wells. The most common method of monitoring fluidcommunication between the reinjection site and the production area is a tracer test (Stefánsson, 1997).This method involves injection of a chemical material (tracer) into the reservoir and measurements oftracer concentration in nearby production wells. A tracer test was designed for the Xi’an geothermalsystem in order to predict potential cooling.A constant mass flow rate, q, is injected down the injection well and a constant mass flow rate, Q, isproduced from the production well, with Q > q. We assume that the flow channel, which connects theinjection well and the production well, is along a narrow fracture zone and the mass flow is, therefore,
- Page 8 and 9: Yin Lihe 422Report 20VARFLOW simula
- Page 13 and 14: Report 20427Yin Lihebreakthrough ti
- Page 15 and 16: Report 20429Yin Lihe5. The effect o
Y<strong>in</strong> Lihe 424Report 20Case II: Four re<strong>in</strong>jection wells,<strong>in</strong>stead <strong>of</strong> two as <strong>in</strong> case I, with aconstant 10 l/s <strong>in</strong>jection rate last<strong>in</strong>g3 months dur<strong>in</strong>g each year’sw<strong>in</strong>tertime. The <strong>in</strong>jection rate is15% <strong>of</strong> the production rate dur<strong>in</strong>gw<strong>in</strong>tertime <strong>of</strong> the year 2000. Figure11 shows the water level recoverycalculated by LUMPFIT.The results <strong>in</strong>dicate that, with 2re<strong>in</strong>jection wells, the water level <strong>in</strong>the system will rise about 7 mmaximum and 2 m m<strong>in</strong>imum with 5m <strong>in</strong> average, long-term. With 4re<strong>in</strong>jection wells, the average waterlevel recovery will be 8 m, and them<strong>in</strong>imum and maximum water levelrecovery will be 5 and 14 m,respectively. There is no doubt thatthe production potential and lifetime<strong>of</strong> the reservoir will improveaccord<strong>in</strong>gly.Water level recovery (m)16128402 re<strong>in</strong>jection wells4 re<strong>in</strong>jection wells2000 2002 2004 2006 2008 2010FIGURE 11: Predicted water level recoverywith 2 and 4 re<strong>in</strong>jection wells4.3 Thermal breakthrough calculationsAll methods, regardless <strong>of</strong> how promis<strong>in</strong>g they may be, have their positive and negative sides. Besidescost consideration, thermal breakthrough is the most serious problem fac<strong>in</strong>g <strong>in</strong>jection. Even thoughthermal breakthrough and cool<strong>in</strong>g <strong>of</strong> reservoir fluid have not been major problems <strong>in</strong> any <strong>geothermal</strong>fields, it is necessary to predict thermal breakthrough time for different <strong>in</strong>jection-production well spac<strong>in</strong>g,i.e. the time from <strong>in</strong>itial <strong>in</strong>jection until a significant cool<strong>in</strong>g is observed <strong>in</strong> a production well (Stefánsson,1997).Consider one re<strong>in</strong>jection well without a production well nearby. The <strong>in</strong>jected water diffuses radially awayfrom the <strong>in</strong>jection well through the porous rock matrix. If an <strong>in</strong>tergranular flow is assumed, then the rockand the fluid have the same temperature at any po<strong>in</strong>t. The differential equation that describes this heattransport is∂Tβw+ q ⋅∇T= 0(7)∂tρβwhere T = Temperature (°C);$ w = Heat capacity <strong>of</strong> water (J/kg/°C); = Wet rock heat capacity (J/°Cm 3 );q = (q x , q y , q z ) the mass flux vector (kg/m 2 /s);LT = The temperature gradient vector.An <strong>in</strong>f<strong>in</strong>ite horizontal reservoir <strong>of</strong> constant thickness H is assumed. Injection <strong>of</strong> Q kg/s <strong>of</strong> cold water isassumed to start at time t = 0. The cold front moves away from the <strong>in</strong>jection well, and the radial distancefrom the well to the temperature front is given as follows:
Change language