CHAPTER 8 Underground Natural Gas Storage 8.1 Introducton In the Unted States and a few other countres, the underground storage of natural gas has become ncreasngly mportant after World War II. The obvous reason for storage s that, tradtonally, natural gas usage has been changng wth seasons. The demand has been hgher n the wnter, prompted by resdental heatng. Thus, the base load and the peak load natural gas, not just n dfferent seasons, but also dfferent days wthn a season, can be qute dfferent. Ths stuaton could create an mbalance between the recepts and delveres of a ppelne network. To avod supply dsruptons, underground storage can be used to provde ppelnes, local dstrbuton companes, producers, and ppelne shppers wth an nventory management tool, seasonal supply backup, and access to natural gas as needed (EIA, 2008). In addton, natural gas storage s also used by ndustry partcpants for commercal purposes: to store gas when gas prce s low and wthdraw and sell gas when the prce s hgh (Speght, 2007). Currently, most of the natural gas storage facltes are n the Unted States, wth very few n Japan and Europe. By the end of 2007, there were about 400 underground storage reservors n the Unted States wth workng gas capacty of ~4,100 Bcf and delverablty rate potental of ~89 Bcf/d (EIA, 2008). There are other ways to store natural gas (such as n lqud form n above-ground tanks as LNG, dscussed n Chapter 6). In ths chapter, we wll only focus on underground natural gas storage. The mpact of LNG on gas storage wll be brefly dscussed at the end of the chapter. 289
290 Chapter 8 Underground Natural Gas Storage 8.2 Types of Underground Storage There are prmarly three types of underground storage facltes, and the descrptons below, wdely acceptable n the ndustry, are taken mostly from the EIA (2004): Depleted ol or gas reservors The advantage of convertng a feld from producton to storage duty s that one can use the exstng wells, gatherng systems, and ppelne connectons. It s usually close to consumpton centers. Ths type of underground storage stes, as shown n Fgure 8 1, s wdely used n the Unted States (about 326 stes, accountng for 82 percent of the total at the begnnng of 2008, EIA, 2008). Aqufers An aqufer s sutable for gas storage f the water bearng sedmentary rock formaton s overlan wth an mpermeable cap rock. Storage s created by njectng gas and dsplacng the water. Therefore, the water movement and cap rock qualty should be taken nto account when selectng and desgnng the storage (Katz and Tek, 1981). Ths type of storage usually requres more base (or cushon) gas (for defnton see Secton 8.3 Storage Measures ) and greater montorng of wthdrawal and njecton performance. Wth the presence of an actve water drve, the delverablty rates may be enhanced. Salt caverns Salt caverns provde very hgh wthdrawal and njecton rates relatve to ther workng gas capacty. Base gas requrements are relatvely low. As shown n Fgure 8 1, the large majorty of salt cavern storage facltes have been developed n salt dome formatons located n the US Gulf Coast States. Salt caverns have also been leached from bedded salt formatons n the Northeastern, Mdwestern, and Southwestern Unted States to take advantage of the hgh njecton/wthdrawal rates and flexble operatons possble wth a cavern faclty. Cavern constructon s more costly than depleted feld conversons when measured on the bass of dollars per thousand cubc feet of workng gas capacty, but the ablty to perform several wthdrawal and njecton cycles each year reduces the per unt cost of each thousand cubc feet of gas njected and wthdrawn. Some recondtoned mne caverns have been n use as well. Hard rock caverns can also be good canddates of gas storage (Heath et al., 1998).
8.3 Storage Measures 291 Fgure 8 1 U.S. Underground natural gas storage facltes n the lower 48 states (EIA 2004) To determne a feld s sutablty as a natural-gas -storage, ts physcal characterstcs such as porosty, permeablty, and retenton capablty should be examned along wth the ste preparaton costs, delverablty rates and cyclng capablty. The good underground storage reservor s obvously the one that has hgh capablty to hold natural gas for future use and hgh delverablty rate at whch gas nventory can be wthdrawn. 8.3 Storage Measures It s necessary to ntroduce some of the concepts used n storage calculaton before we go to the detaled calculaton of the storage capacty. For consstency, here we use the same defntons as they are used by EIA (2004): Total gas storage capacty the maxmum volume of gas that can be stored n an underground storage faclty by desgn. It s determned by the physcal characterstcs of the reservor and nstalled equpment. Total gas volume n storage the volume of storage n the underground faclty at a partcular tme.
292 Chapter 8 Underground Natural Gas Storage Base gas or cushon gas the volume of gas ntended as permanent nventory n a storage reservor to mantan adequate pressure and delverablty rates throughout the wthdrawal season. It contans two elements (Tureyen et al., 2000): Recoverable base gas the porton of the gas that can be wthdrawn wth current technology, but t s left n the reservor to mantan the pressure. Non-recoverable base gas the porton of the gas that cannot be wthdrawn wth the exstng facltes both techncally and economcally. The relatonshp among the total gas storage capacty, total gas volume n storage, and base gas s llustrated n Fgure 8 2. Workng gas capacty the total gas storage capacty mnus base gas,.e., the volume of gas n the reservor above the level of base gas. So, for a gven storage capacty, the hgher the base gas s, the lower the workng gas wll be, the less effcent the storage wll be. Injecton volume the volume of gas njected nto storage felds durng a gven perod. Delverablty or delverablty rate, wthdrawal rate, wthdrawal capacty a measure of the amount of gas that can be delvered or wthdrawn from a storage faclty on a daly bass wth the unt of MMscf/d, same as that for producton rate. Occasonally, t s expressed n terms of equvalent heat content of the gas wthdrawn from the faclty such as dekatherms per day. A therm s roughly equvalent to 100 scf of natural gas; a dekatherm s about 1 Mscf. In general, a faclty's delverablty rate vares drectly wth the total amount of gas n the reservor; t s at ts hghest when the reservor s most full and declnes as workng gas s wthdrawn. Injecton capacty or rate the amount of gas that can be njected nto a storage faclty on a daly bass. As wth delverablty, njecton capacty s usually expressed n MMscf per day, although dekatherms per day s also used. By contrast, the njecton rate vares nversely wth the total amount of gas n storage; t s at ts lowest when the reservor s most full and ncreases as workng gas s wthdrawn.
8.3 Storage Measures 293 Storage Measures, Bcf Total Storage Capacty Total Gas n Storage Base Gas Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Tme Fgure 8 2 Storage measures These measures for any gven storage faclty are not necessarly absolute and are subject to change or nterpretaton. In the followng sectons, natural gas storage s vewed n terms of a depletng or ncreasng pressure n a closed reservor wthout actve water drve. If the reservor pressure s supported by actve water movement, equatons have to be modfed (Katz and Tek, 1981; Mayfeld, 1981). 8.3.1 Total Gas Volume and Injected Gas Volume n Storage The njected gas volume n a depleted gas reservor can be calculated by usng a smlar approach as dscussed n Secton 1.6.4 Gas Formaton Volume Factor of Chapter 1 for the ntal gas-n-place calculaton of a producng feld (Eq. (1.13)). Assume the reservor pore volume s constant, the ntal gas-n-place n the depleted gas reservor n standard condtons s G, and the total gas volume n storage faclty s G, then the cumulatve njected gas volume, G s s G G- G, s (8.1) or, by employng the formaton volume factors at ntal and fnal condtons G G B Ê G G B ˆ g g - - B Á 1. Ë B g g (8.2)
294 Chapter 8 Underground Natural Gas Storage Note: the G s the resdual gas n a depleted gas reservor that wll be used for storage, or the ntal gas n a storage feld after the seasonal wthdrawal and at the begnnng of the resumpton of njecton. It can be calculated by usng Eq. (1.13). Substtutng Eq. (1.12) nto Eq. (8.2) and assumng the temperature s constant, Eq. (8.2) becomes Ê G G pz ˆ G p p s - Ë Á p Z Ê p Z Z - ˆ 1 Ë Á Z. (8.3) In Eqs. (8.1 to 8.3), the subscrpt stands for the ntal condtons of the gas storage. The pressures are measured when the storage s at ts maxmum and mnmum capactes. The pressures measured are then near the maxmum and mnmum pressures. Eq. (8.3) s vald when there s no actve water drve. Example 8 1 Calculaton of total gas volume A depleted gas reservor s converted to natural gas storage. The reservor data and condtons are gven n Table 8 1. Calculate the total gas volume n the reservor and the total njected gas volume at p 6,000 ps. For convenence, Z s gven as 1.07 (otherwse t can be calculated by usng the correlatons gven n Chapter 1 wth g g 0.6). Assume the temperature wll be the same as the ntal temperature. Table 8 1 Input Parameters for Example 8 1 Varable Quantty Unt A 200 acre h 50 ft f 0.25 S w 0.25 g g 0.6 T 150 F p 1,000 ps Z 0.91
8.3 Storage Measures 295 Soluton Use Eq. (1.12) for the calculaton of the formaton volume factors 0.91 (150 + 460) B g 0.0283 1,000 1.07 (150 + 460) B g 0.0283 6,000 0. 0157 3 res ft /scf, 3 0. 0031 res ft /scf. Use Eq. (1.13), at 1,000 ps 200 50 0. 25 ( 1-0. 25) G 43, 560 5, 202 MMscf. 0. 0157 1, 000, 000 Total gas volume n storage at 6,000 ps can be calculated as G 5, 202 0. 0157 26, 346 MMscf. 0. 0031 The cumulatve gas volume njected can be obtaned from Eq. (8.1) or by usng Eq. (8.2) G s 26, 346-5, 202 21, 144 MMscf, Ê ˆ G s 5, 0. 0157 202 - Ë Á 1 21, 144 MMscf. 0. 0031 Ths s an mportant exercse as, n realty, the ntal gas-n-place for a gven storage s often not known. By recordng the cumulatve njected gas volume at gven condtons (p and T) and assumng the temperature s constant at all tme (a reasonable assumpton), then p/z versus G s can be plotted. If there s no aqufer support, ths lne should be straght, as demonstrated n Fgure 8 3, and the slope can be determned. Rearrangng Eq. (8.3) gves p Z Gs( p / Z) p +. G Z (8.4)
296 Chapter 8 Underground Natural Gas Storage p/z, ps Slope(p /z )/G p /z G s, Bcf Fgure 8 3 p/z curve vs cumulatve gas storage A plot of p/z versus G s should yeld a straght lne and the slope should be (p /Z )/G. Therefore the ntal gas-n-place can be obtaned by G ( p Z)/ slope. (8.5) p /Z can be determned by measurng the pressure at ntal condtons through a pressure buldup test. Example 8 2 Calculaton of ntal gas-n-place Determne the ntal gas-n-place for a shallow, low pressure gas storage reservor. The njected gas over tme and the p/z data are gven n Table 8 2. Table 8 2 Input Data for Example 8 2 Year Season G s, Bcf p/z, psa Year Year + 1 Sprng 13.5 365 Fall 17.6 470 Sprng 14.5 389 Fall 17.5 465
8.3 Storage Measures 297 p/z, psa 470 450 430 410 390 370 350 13 14 15 16 17 18 G s, Bcf Year Year +1 Fgure 8 4 p/z vs gas storage for Example 8 2 Soluton Plot p/z versus G s (see Fgure 8 4) by usng the data provded n Table 8 2. Obvously ths s an deal case as t shows the slopes from both Year and Year + 1 are pretty much the same and s about 25.5 psa/bcf. Extrapolate the lne and ntercept t wth the vertcal axs. Ths gves p /Z 21.0 psa (at G s 0). Use Eq. (8.5), the ntal gas-n-place for ths gven gas storage s G 21. 0 / 25. 5 0. 824 Bcf. Ths s also a good tool to evaluate the gas losses n storage, whch s one of the crtcal ssues n gas storage that should be addressed. 8.3.2 Losses n Gas Storage 1 Gas loss n gas storage s a very serous ssue. It happens when the cap rock does not seal well, cement around the wellbore s flawed, or there s a communcaton between the storage and other reservors. Once gas loss s happenng, the storage delverablty or wthdrawal rate wll declne from year to year, and the operator wll have to bear wth hgh cost or even the rsk of not meetng the peak demand. A 1. Some of the materal n ths secton s contrbuted by Phl Lews, 2009.
298 Chapter 8 Underground Natural Gas Storage report (Neukarn, 2008) showed that the annual losses can be up to 0.5 Bcf. If the gas prce s $4/Mscf that means ths storage s losng $2 mllon per year, whch s a sgnfcant loss. Therefore, gas storage must be montored properly to determne the magntude of such loss, the root cause, and remedy t as soon as t s detected. For gas storage that s converted from depleted gas reservor wth no water drve, the gas flows to the wells prmarly by gas expanson. Then a procedure can be used to determne the gas loss (Mayfeld, 1981). There are several ways to determne the reservor pressure. One way s to conduct regular (e.g., semannual) pressure buld-up tests smlar to pressure surveys done n gas producton felds. Another way s to montor the bottomhole pressure n observaton wells. Ordnarly, these pressure surveys are conducted n the fall and sprng when reservor pressure s near maxmum and mnmum for total gas volume calculaton (as dscussed n Secton 8.3.1 Total Gas Volume and Injected Gas Volume n Storage ). The preferred observaton well s the one at the locaton wth the hghest permeablty. The plot s usually smoother and more relable for the njecton season as the njecton rate s usually constant. Durng the wthdrawal season, fluctuaton can happen as the demands from ppelne systems can be dfferent (Mayfeld, 1981). The total gas n storage or gas-n-place can be plotted along wth the determned p/z. If there s no gas loss, all data ponts should fall on the same lne after repeated cycles of njecton and wthdrawl. If the slope of the lne becomes smaller, ths s lkely to mean that the storage ncreases because of gas mgraton or leakage. When there s gas loss, parallel lnes would appear from year to year and are shfted towards a larger gas volume at a gven p/z. The dfference between these lnes s gas loss. Ths can be seen n Example 8 3. Example 8 3 Calculaton of gas loss Assume ths s the same storage reservor as that shown n Example 8 2. After a few years, well delveres started declnng. The bottomhole pressure over Z and gas njected n Year + 2 are collected and summarzed n Table 8 3. Soluton Plot p/z versus G s for dfferent years n Fgure 8 5. Results show that the lne from Year s overlan wth that from Year + 1. The lne from Year + 2 s parallel wth those from Year and Year + 1 but shfted towards a larger G s. Ths mples that the storage s losng gas.
8.3 Storage Measures 299 Table 8 3 Data for Example 8 3 Year Season G s, Bcf p/z, psa Year Sprng 13.5 365 Fall 17.6 470 Year + 1 Sprng 14.5 389 Fall 17.5 465 Year + 2 Sprng 15.1 395 Fall 17.7 460 480 460 p/z, psa 440 420 400 380 Year Year +1 Year +2 360 13 14 15 16 17 18 G s, Bcf Fgure 8 5 p/z versus G s plot for Example 8 3 From the data set of Year and Year + 1, Eq. (8.4) yelds Ê p ˆ Ë Á Z 25. 5 G 21. 0. ( s ) + + 1 + 1 Smlarly, from the data set of Year + 2, Eq. (8.4) yelds Ê p ˆ Ë Á Z 25. 2 G 4. 4. ( s ) + + 2 + 2
300 Chapter 8 Underground Natural Gas Storage Choose p/z 465 psa, then (G s ) +1 17.4 Bcf, and (G s ) +2 17.9Bcf. So the gas loss (G s ) +2 (G s ) 0.5 Bcf. Gas loss can also be determned by plottng G s /(p/z) versus tme (year). If G s /(p/z) does not change wth tme, t s an ndcaton that the storage faclty s secure. If the values are ncreased wth tme, that wll be an ndcaton that ether the storage s losng gas or the effectve sze of the storage s ncreased. The amount of gas lost can be determned by usng the procedure outlned above. 8.3.3 Injectvty n Gas Storage Well The expresson for njectvty of a gas storage well can be nferred from the expressons for the productvty of a gas well, rememberng that n storage, gas s njected nto a closed system (unless there s a leak). So steady state s not applcable n njectvty evaluaton of gas storage wells. Under pseudosteady state, the njectvty can be calculated by q nj kh( pnj - p ). (8.6) 0 re, ZT[ln(. 472 1 424m ) + s] r 2 2 For transent flow, n terms of real gas pseudopressure, w q nj khmp [ ( nj )- mp ( )] È k Ílogt + log - 323. + 087. s 1, 638T ( ct ) 2 Î fm rw or, n terms of pressure squared dfference, -1, (8.7) q nj 2 2 kh[ pnj - p ] È k Ílogt + log - 323. + 087. s, ZT Î ( ct ) 2. 1 638m fm rw -1 (8.8) In Eq. (8.7), the m(p) s defned n Eq. (3.19). In Chapter 3, we presented a comprehensve method of combnng materal balance (p/z versus G p ) along wth well delverablty, and showed how to establsh a forecast of well performance. The producton rate decreases as the reservor pressure decreases. In storage, the njecton rate may also decrease as the reservor pressure
8.4 Dscusson 301 ncreases, therefore the drvng pressure dfference decreases for a constant njecton pressure. Example 8 4 Calculate the njecton rate of a well n a gven gas storage Gven: the well bottomhole njecton pressure s 3,000 ps. The reservor pressure at the tme and the temperature are 1,500 ps and 200 F, respectvely. r e 660 ft, r w 0.359 ft, k 1 md, and h 45 ft. The average Z-factor and vscosty are 0.897 and 0.0175, respectvely. Repeat the calculaton when the reservor pressure s 2,000 ps. (The average Z-factor and vscosty are 0.890 and 0.0181 cp, respectvely). Soluton Use Eq. (8.6), 2 2 1 45 ( 3, 000-1, 500 ) q nj 0. 472 660 1, 424 0. 0175 0. 897 ( 200 + 460) [ln( ) + 0] 0. 359 3, 040 Mscf/d. Repeatng the above calculaton for average storage pressure equal to 2,000 ps, the njecton rate s 2,200 Mscf/d, showng the mpact of the pressurzaton of the reservor on well njectvty. 8.4 Dscusson The emergence of LNG as a major contrbutor to natural gas supply n the Unted States wll most certanly alter tradtonal storage patterns and ther seasonalty. Whle the calculatons presented n ths chapter wll stll be vald, n practce, there wll probably be a lot fewer large cycles, such as one n the summer and one n the wnter, of storage njecton and producton as has been the case n the past. Instead cycles may be a lot smaller and repeated several tmes n a year; reflectng weather nduced hgh and low demand of heatng or ar condtonng loads. Management of gas storage, wth ts ablty to nject and wthdraw relatvely quckly n conjuncton wth a steady or dscreet supply of LNG, becomes an mportant new dmenson n natural gas use.
302 Chapter 8 Underground Natural Gas Storage 8.5 References EIA. 2004. The bascs of underground natural gas storage. Natural Gas Dvson. EIA. 2008. Underground natural gas storage. Heath, S.M., R.C. Hodren, E. Kostaks, J.P. Harrson. 1998. Underground storage of natural gas n unlned hard rock caverns. Paper SPE 47221. Katz, D.L. and M.R. Tek. 1981. Overvew on underground storage of natural gas. JPT 33 (6). Mayfeld, J.F. 1981. Inventory verfcaton of gas storage felds. JPT 33 (9). Neukarn, J. Response to Federal Energy Regulatory Commsson s Data Requests, Southern Star Central Gas Ppelne, Inc., Docket No CP08-4- 000. Dated February 8, 2008. Speght, J.G. 2007. Natural Gas: A Basc Handbook. Houston: Gulf Publshng Company. Tureyen, O.I., H. Karaaloglu, and A. Satman. 2000. Effect of the wellbore condtons on the performance of underground gas-storage reservor. Paper SPE 59737.