Geophysical Journal InternationalGeophys. J. Int. (2013) 193, 1370–1384 doi: 10.1093/gji/ggt015Advance Access publication 2013 April 01

GJI

Mar

ine

geos

cien

ces

and

appl

ied

geop

hysics

Invasion of drilling mud into gas-hydrate-bearing sediments. Part I:effect of drilling mud properties

Fulong Ning,1,2 Keni Zhang,3,4 Nengyou Wu,2 Ling Zhang,1 Gang Li,2 Guosheng Jiang,1

Yibing Yu,1 Li Liu1 and Yinghong Qin1

1Faculty of Engineering, China University of Geosciences, Wuhan, Hubei 430074, China. E-mail: nflzx770803@163.com2Guangzhou Center for Gas Hydrate Research, Chinese Academy of Sciences, Guangzhou 510640, China3Earth System Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA4College of Water Sciences, Beijing Normal University, Beijing, 100875, China

Accepted 2013 January 16. Received 2012 November 30; in original form 2011 November 12

S U M M A R YTo our knowledge, this study is the first to perform a numerical simulation and analysis ofthe dynamic behaviour of drilling mud invasion into oceanic gas-hydrate-bearing sediment(GHBS) and to consider the effects of such an invasion on borehole stability and the reliabilityof well logging. As a case study, the simulation background sets up the conditions of mudtemperature over hydrate equilibrium temperature and overbalanced drilling, considering thefirst Chinese expedition to drill gas hydrate (GMGS-1). The results show that dissociating gasmay form secondary hydrates in the sediment around borehole by the combined effects ofincreased pore pressure (caused by mud invasion and flow resistance), endothermic coolingthat accompanies hydrate dissociation compounded by the Joule–Thompson effect and thelagged effect of heat transfer in sediments. The secondary hydrate ring around the boreholemay be more highly saturated than the in situ sediment. Mud invasion in GHBS is a dynamicprocess of thermal, fluid (mud invasion), chemical (hydrate dissociation and reformation) andmechanical couplings. All of these factors interact and influence the pore pressure, flow ability,saturation of fluid and hydrates, mechanical parameters and electrical properties of sedimentsaround the borehole, thereby having a strong effect on borehole stability and the results of welllogging. The effect is particularly clear in the borehole SH7 of GMGS-1 project. The boreholecollapse and resistivity distortion were observed during practical drilling and wireline loggingoperations in borehole SH7 of the GMGS-1.mud density (i.e. the corresponding boreholepressure), temperature and salinity have a marked influence on the dynamics of mud invasionand on hydrate stability. Therefore, perhaps well-logging distortion caused by mud invasion,hydrate dissociation and reformation should be considered for identifying and evaluating gashydrate reservoirs. And some suitable drilling measurements need to be adopted to reduce therisk of well-logging distortion and borehole instability.

Key words: Downhole methods; Geomechanics; Gas and hydrate systems; Ocean drilling;Phase transitions.

1 I N T RO D U C T I O N

Gas hydrates are non-stoichiometric inclusion compounds formedwhen hydrophobic gas molecules (usually methane and carbondioxide) come into contact with water (host molecules) under con-ditions of low temperature (typically <300 K) and high pressure(typically >3.8 MPa at a temperature of 277 K; Sloan 2003). Gashydrates are widely distributed in marine continental margin sed-iments and to a lesser degree in permafrost environments (Kven-volden 1993).

Great advances have been made in the exploration for and ex-ploitation of natural gas hydrates (Jaiswal et al. 2006; Priest et al.2006; Weitemeyer et al. 2011). Yet further development is hamperedby poor quantitative descriptions of in situ properties of hydrate de-posits, immature exploration and production technology and highsafety risk of geology and environment in exploitation. The prob-lems of immature technology and safety risk are related to factorssuch as drilling safety (e.g. wellbore instability), formation defor-mation, geologic hazards and climate change. Especially for poorlyconsolidated hydrate deposits in marine environments (Birchwood

1370 C©The Authors 2013. Published by Oxford University Press on behalf of The Royal Astronomical Society.

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1371

Figure 1. Schematic diagram of mud invasion into gas-hydrate-bearing sediment. Left-hand side: in the case of over-balanced drilling conditions, mud invadesthe GHBS. Right-hand side: pore-scale schematic diagram of mud invasion accompanying hydrate dissociation.

et al. 2008), drilling induced wellbore instability is likely to occurif underbalanced drilling is adopted (Collett et al. 2009; Birchwood& Noeth 2012). Under such drilling conditions, gas hydrates tendto decompose because of the pressure decrease and further causesa sharp reduction in the strength of the sediment (Winters et al.1999a,b, 2007), thereby placing the wellbore at risk of collapse(Freij-Ayoub et al. 2007). A practical way to reduce this risk is tomaintain the wellbore pressure at a higher level than the pore pres-sure but lower than the formation fracture-opening pressure (Ninget al. 2008a). In such a case, mud (herein referred to as ‘water-basedmud’) displaces the original pore fluid surrounding the borehole andinvades the gas-hydrate-bearing sediment (abbreviated as GHBS)because of the pressure gradient between the mud and the surround-ing settlement (Fig. 1). Mud invasion into drillholes has an effect onhost-sediment properties, such as permeability, pore pressure andmechanical strength (Phelps 1995; Chi et al. 2006), and it is alsolikely to influence the stability of gas hydrates. For example, thehydrate may dissociate due to the relative high temperature of themud and due to frictional heat generated by the drilling tool.

In addition to its effect on wellbore stability, hydrate dissociationhas a strong influence on geophysical well logging, especially theresistivity and wave velocity of sediments. That is because the re-sistivity and wave velocity of sediments are influenced greatly byoccurrence of gas hydrates among numerous geophysical loggingproperties (Collett & Lee 2000; Collett 2001; Spangenberg 2001;Chand et al. 2006; Riedel et al. 2006; Waite et al. 2009). In marineareas, the acoustic logging method is more sensitive to formationconsolidation than other logging methods, suggesting that resistivitywell logging is more reliable than acoustic logging because gas hy-drates normally occur in relatively unconsolidated sediments (Lee &Collett 2008; Lee & Collett 2009b). However, a water-based drillingmud system, containing polymer and specified concentrations ofsalts, is commonly used when drilling in deep water (Ebeltoft et al.1997) or when drilling marine hydrate deposits (Collett et al. 2009;Jiang et al. 2011). In addition to borehole washout caused by thistype of drilling mud circulation (Lee et al. 2012), mud-filtrate inva-sion may also have a strong effect on the reliability of geophysicalresistivity logging. Salts are thermodynamic inhibitors, and whencarried with invading mud, they potentially cause hydrate dissocia-tion (Dickens & QuinbyHunt 1997), leading to wellbore instabilityand affecting logging data. Therefore, it is important to understandthe dynamics of drilling mud invasion and to consider its influenceon the GHBS, especially in terms of the response of geophysicalwell logging in the GHBS, the evaluation of reservoirs, boreholestability, regional hydrate resources and environmental assessment.

In the case of overbalanced conditions, the temperature and salin-ity of the drilling mud are major factors in determining hydrate sta-bility in the sediments during mud invasion. If the temperature ofthe mud falls below the hydrate equilibrium temperature (Point A in

Figure 2. Relation between mud temperature and the temperature of GHBS.There are three possible cases when drilling in GHBS: drilling mud tem-perature is lower (point A), equal to or higher (point B) than the hydrateequilibrium temperature in in situ sediment. We focused on the last case.

Fig. 2) under given salinity and pressure conditions, the hydrate re-mains undissociated; consequently, the mud invasion influences theGHBS in a manner similar to the case of conventional oil/gas reser-voirs. Compared with permafrost hydrate deposits, the temperatureof an oceanic GHBS is higher and closer to the phase equilibriumline (Booth et al. 1996). For example, in the Shenhu area of theSouth China Sea, the temperature and pressure of GHBS are closeto those of phase equilibrium (Ning et al. 2010). In such a case,borehole drilling may easily lead to hydrate dissociation around theborehole because drilling generated heat, salts as thermodynamicinhibitors in mud and originally higher mud temperature togethermay jointly result in the local temperature of mud crossing thephase equilibrium boundary (Point B in Fig. 2). Such conditionswere encountered when drilling the WR 313-G well during the Gulfof Mexico Gas Hydrate Joint Industry Project Leg II (JIP Leg II)in May 2009 (Collett et al. 2009). Under this condition, the overallbehaviour of mud invading, the GHBS can be described as a multi-phase flow coupled with hydrate dissociation driven by differentialpressure and differential temperature. Existing numerical models ofmultiphase fluid-flow for hydrate production may therefore be usedto investigate the invasion process of drilling mud in the GHBS.The difference between gas production from the GHBS and mudinvasion into the GHBS is that the influence range of mud invasionis small, limited to an annular area around the borehole, contrarydirection of multiphase flow and hydrate dissociation occurs closeto the borehole.

On the basis of a previous theoretical analysis (Ning et al.2008b, 2009), we performed a series of numerical simulations toinvestigate the characteristics of drilling mud invasion and its in-fluence on hydrate sediments around a borehole. This simulation

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1372 F. Ning et al.

considers the case of Project GMGS-1 in the South China Sea, us-ing the TOUGH+HYDRATE code developed by Lawrence Berke-ley National Laboratory (Moridis et al. 2009a). The effect of mudinvasion on GHBS can be discussed from two aspects: the proper-ties of drilling mud and the petrophysics of sediment. This paperconsiders the former, with the aim of identifying the invasion char-acteristics of drilling mud into GHBS and the key factors that affectthe mud invasion. The results of this study may help in guidingwell drilling, logging and the protection of hydrate reservoirs, espe-cially correcting for the distortion of well logging induced by mudinvasion.

2 S I M U L AT I O N M E T H O D

2.1 Background

The first Chinese expedition to drill gas hydrate, GMGS-1, wascarried out in the Shenhu area of the South China Sea during 2007April to 2007 June by the companies Fugro and Geotek for theGuangzhou Marine Geological Survey (GMGS), the China Geo-logical Survey (CGS) and the Ministry of Land and Resources ofP. R. China (Fig. 3). Eight sites were selected for drilling and welllogging. Cores were recovered at five sites, among which gas hy-drate occurrences were identified from Site SH2, SH3 and SH7(Wu et al. 2009). In the three sites, hydrates occur at 153–229 mbeneath the seafloor, with a thickness of 10–43 m and porosity of33–48 per cent. The hydrate is mainly type I methane hydrate with asaturation of 26–48 per cent (estimated from well logging), and it isdisseminated throughout the sediment. The produced gas consists of96.1–99.82 per cent methane, which was of microorganism origin.In situ measurements indicate that the bottom-water temperaturewas 3.3–3.7◦C and the geothermal gradient was 43.0–67.7◦C km–1,corresponding to a sea-bottom heat flow of 74.0–78.0 mW m–2

(average 76.2 mW m–2).

During the GMGS-1 Project, drilling was performed from thevessel Bavenit (of the company AMIGE), which does not have a ma-rine riser. The assembly for well drilling was as follows: Ø228.6 mmblade core bit + long-sealing pup joint + short-sealing pup joint +landing nipple + shut-off nipple + floating valve nipplet + Ø177.8-mm drill collar (×8) + Ø127-mm drilling pipe (×22) + Ø175-mmaluminium drilling pipe (×100) + Ø127-mm drilling pipe. Sea wa-ter was used as drilling mud with a rate of 70–80 rpm and displace-ment of 800–900 L/min. When the well reached the desired depth,sea water in the open-hole annulus was displaced by bentonite mudwith a density of 1190 kg m–3 to improve wellbore stability. Welllogging and coring were conducted to characterize the borehole (Huet al. 2009). Mud invasion into the GHBS mainly occurred duringthe stage of mud displacement. During this stage, mud circulationoccurred for only a short time, meaning that the mudcake effect didnot necessarily need to be considered.

The software TOUGH+HYDRATE is widely used to simulategas recovery from hydrate reservoirs in marine or permafrost re-gions (Moridis et al. 2007, 2009b; Li et al. 2010), and it can be usedin tandem with other software (e.g. FLAC 3D) to simulate wellborestability and sediment deformation during gas production from hy-drate deposits (Rutqvist et al. 2008, 2009). TOUGH+HYDRATEcan simulate non-isothermal hydration reactions, phase behaviourand the flow of fluids and heat under the conditions typical of naturalCH4-hydrate deposits in complex geologic media. It includes bothequilibrium and kinetic models of hydrate formation and dissocia-tion. The model accounts for heat and up to four mass components(i.e. water, methane gas, hydrate and water-soluble inhibitors suchas salts or alcohols) that are partitioned among four possible phases:gas, aqueous liquid, ice and hydrate (Moridis et al. 2009a). Up to15 states (phase combinations) can be described by the code, andit is capable of simulating any combination of hydrate dissociationand of describing phase changes and heat mass transfer for sedi-ment fluids, as well as considering other typical hydrate problems(Zhang et al. 2008). We adopted the equilibrium model of hydrate

Figure 3. (a) Location of the Shenhu GMGS-1 project on the north slope of the South China Sea. (b) Bathymetric map of the area of gas hydrate drilling,showing the locations of wells drilled in the research area (reproduced from Li et al. 2010).

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1373

formation and dissociation, and did not consider chemical and me-chanical coupling of the diffusion effect. For simplification, weassumed that the mud only contained the thermodynamic inhibitorNaCl.

2.2 Simulation model and parameters

The simulations were based on data retrieved from borehole SH7,where the seafloor is at a water depth of 1108 m. The GHBSmainly occurred in fine-grained sediment at 155–177 m below theseafloor (mbsf), including sandy clay at drillhole depths of 155.0–159.4 mbsf, silty clay at 159.4–168.9 mbsf and firm olive-grey clayat 168.9–177 mbsf. The hydrate concentration at the site is variable,with a maximum value of 44 per cent of pore volume calculatedfrom porewater freshening. Here 44 per cent is selected as the orig-inal hydrate saturation for the simulations. A thickness of 0.1 m inthe middle of the GHBS (about 166 mbsf) is investigated. In thisthin layer, the GHBS is regarded as isotropic sediment in which hy-drates occur as pore filling habit. The methane hydrates and watercoexist in the in situ sediment. An axisymmetric cylinder with aradius of 5 m is adopted for the model domain, with the boreholelocated in the centre of the cylinder. The investigated thin layer wasdiscretized into 100 radial grid cells. The drilling tools assemblyused a borehole with a diameter of 228.6 mm, a drilling pipe with adiameter of 177.8 mm and an annular with a clearance of’ 25.4 mm.The mud in the borehole is treated as a fixed inner boundary; thatis, the temperature and pressure remained stable in this thin mudlayer. The process of mud invasion is simplified as a 1-D radialdisplacement problem. The physical model is shown schematicallyin Fig. 4.

Figure 4. Schematic mud-invasion model at SH 7 site of the GMGS-1project. Red arrows indicate the direction of drilling mud flow. The hole wasdrilled without the use of a riser. The drilling mud was directly dischargedinto sea water through the ‘open hole’. The dashed rectangle in the gas-hydrate-bearing layer represents a thin layer located in the centre of thehydrate-bearing layer, which was selected as research area and divided into100 grid cells. Mud invasion in this thin layer was simplified as a 1-D radialdisplacement problem.

Because natural gas hydrates are mainly distributed in poorlyconsolidated sediments near the seafloor, pore water in the sed-iment can be considered to communicate with sea-bottom water.Then the hydrostatic pore-water pressure can be calculated on ba-sis of well depth and average sea water density. The geothermalgradient at borehole SH7 was 43.2◦C km–1 and the sea-bottom tem-perature at the site was about 3.5◦C (Li et al. 2010). The tempera-ture and pressure at the thin layer were calculated to be 13.74◦Cand 13.2 MPa, respectively, which is within the stability fieldof hydrate.

The drilling mud temperature at the simulation depth was es-timated to be close to 15◦C, based on the borehole conditions oflogging at the depth of hydrate occurrence (Nakai et al. 2007),and mud pressure was estimated from the pressure of overlying seawater and mud in the borehole, as follows:

Pf = Patm + g(ρswh + ρf z) × 10−6, (1)

where Pf is mud pressure, MPa; and ρf is mud density, kg m–3;Patm is standard atmospheric pressure, 0.101325 MPa; h is waterdepth, m; z is the depth of sediment from the sea bottom, m; g is theacceleration of gravity, m s–2 and ρsw is the average sea water density,which is a function of water depth, temperature and salinity, kg m–3.The calculated Pf was about 13.54 MPa, at which the correspondingequilibrium temperature was about 14.7◦C for a NaCl concentrationof 3.05 per cent, suggesting that the mud temperature crosses the gashydrate stability phase boundary. The model parameters are listedin Table 1.

3 R E S U LT S A N D A NA LY S I S E S

3.1 Characteristics of drilling mud invasion into GHBS

The simulated distributions of pressure and temperature aroundthe borehole during mud invasion are shown in Figs 5a and b.During drilling through the GHBS, mud filters moved swiftlyinto the borehole wall and displaced the original fluids, causinga rapid pressure increase in the local sediment. With the ongo-ing infiltration of mud, the zone of enhanced temperature andpressure gradually extended farther into the surrounding sedi-ment. The temperature spread, which reached 0.5 m from theborehole at 5 h and 1.0 m at 24 h, lagged behind the pressurespread (Figs 5a and b), which reached 5 m from the borehole at5 h. During mud invasion, the temperature and pressure devel-oped in a similar manner to that in conventional oil and/or gassediment.

Despite this similarity, during hydrate drilling the mud invasionwas accompanied by phase changes, as indicated by variations inthe saturation of the hydrate, aqueous and gas phases around theborehole. This differs from mud invasion in conventional sedimentand GHBS. For example, in the present simulation, hydrate dis-sociation occurred during mud invasion and led to a reduction insaturation, concomitant with an increase in water and free-gas satu-ration around the borehole (Fig. 5c). The increase in water contentis due to mud invasion and hydrate dissociation. Fig. 5c also showsthat the mud invasion causes hydrate dissociation accompanied byhydrate reformation under the simulation conditions. Such reformedhydrates are called ‘secondary hydrates’ (Moridis & Reagan 2007),and these have been predicted by numerical simulations (Moridis &Reagan 2007; Li et al. 2010) and validated in experiments (Kneafseyet al. 2007; Seol & Myshakin 2011). The reformation of secondaryhydrates occurs because of the combined effect of several factors:

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1374 F. Ning et al.

Table 1. Main properties and conditions of mud and hydrate deposits at Site SH7.

Parameter Value Parameter Value

Initial temperature of GHBS 13.74◦C Grain density (ρs) 2600 kg m–3

Initial pressure of GHBS 13.2 MPa Grain specific heat (cs) 1000 J kg−1 ◦C−1

Temperature at sea bottom 3.5◦C Geothermal gradient 0.0432◦C m–1

Pressure at the sea bottom 11.34 MPa Wet thermal conductivity (λs) 3.1 W m−1 ◦C−1

Initial water salinity (mass fraction) 3.05 per cent Dry thermal conductivity (λHs) 0.85 W m−1 ◦C−1

Hydrate saturation in situ (sH) 0.44 Thermal conductivity of hydrate (λH) 0.5 W m−1 ◦C−1

Pore-water saturation in situ (sA) 0.56 Intrinsic permeability (k) 7.5 × 10−14 m2 ( = 75 mD)Mud temperature 15◦C Hydrate density (ρH) 920 kg m–3

Mud density (ρf) 1190 kg m–3 Specific heat of hydrate (cH) 2100 J kg−1 ◦C−1

Porosity (�) 0.40 Specific heat of mud 4000 J kg−1 ◦C−1

Compression coefficient 1.00 × 10–8 Pa−1 Gas composition 100 per cent CH4

Composite thermal conductivity modela λc = λHs + (S1/2A + S1/2

H )(λs − λHs) + φSIλI, here SI = 0Capillary pressure modelb Pcap = −Ps[(S∗)−1/λ − 1]1−λ, S∗ = (SA − S′

irA)/(SmxA − S′irA),

−Pmax ≤ Pcap ≤ 0, λ = 0.45, SirA’ = 0.29, SmxA = 1.0, Pmax = 105 Pa

Relative permeability modelckrA = [(SA − SirA)/(1 − SirA)]n, krG = [(SG − SirG)/(1 − SirA)]nG

, krH = 0SirA = 0.30, SirG = 0.05, n = nG = 3.572.

aMoridis et al. (2005, 2009a).bvan Genuchten (1980).cStone (1970).

the squeezing effect of drilling mud invasion (i.e. increasing porepressure due to mud invasion and flow resistance), endothermiccooling related to hydrate dissociation compounded by the Joule–Thompson effect (Moridis & Reagan 2007) and the lagged effect ofheat transfer in sediments. Because of these factors, the local hydratethermodynamic parameters fall below the phase equilibrium bound-ary. The relationship between mud invasion and hydrate reformationis, to some extent, similar to injecting the constituent componentsof hydrate (i.e. water and methane) for permeability measurementsof GHBS in displacement experiments, which would likely leadto additional hydrate formation (Johnson et al. 2011). In addition,water from hydrate dissociation is displaced by mud invasion anddilutes the local salinity of pore water (Fig. 5d), further promotingthe reformation of gas hydrate (Li et al. 2010; Seol & Myshakin2011).

The occurrence of secondary hydrate, together with existing hy-drate, means that the hydrate saturation is locally even higher thanthe original hydrate saturation (0.44). In the simulations, such ahigh-saturation hydrate ring formed, with a peak value of 0.4617 ata radial distance of 0.15 m from the borehole after about 12 h. Theoccurrence of this ring and free gas led to a decrease in local watersaturation, forming ‘water troughs’ (Fig. 5c). The high-saturationhydrate ring may be attributed to the displacement effect of mud in-vasion and to permeability hysteresis related to secondary hydrates,which causes a reduction in permeability (Konno et al. 2008). Thus,the dissociated gas and water accumulate in this area and ultimatelyform the ring. The simulation results (Table 2) also show thereoccurs in situ hydrate dissociation following the high-saturation hy-drate ring, although this is not apparent in Fig. 5c (this is furtherdiscussed in Section 4.1). For example, at 24 h, the hydrate satu-ration was 0.4612 at 0.17 m from the borehole, but just 0.4388 at0.22 m from the borehole, which is lower than the original satu-ration of 0.44. Therefore, the high-saturation hydrate ring inducedby secondary hydrates can serve to divide the whole hydrate disso-ciation area into two parts: from the borehole face to the positionof the ring, and outward from the ring, with the degree of hydratedissociation much less in the outward part. The aforementionedprocess of mud invasion accompanying with hydrate dissociationand high-saturation hydrate ring is described in Fig. 6 and wasvalidated indirectly from the 20 and 24 h curves in Fig. 5d. The

reformation of hydrate resulted in an increase in pore-water salin-ity while weak hydrate dissociation resulted in a salinity decrease(Fig. 5d). The mud invasion and heat transfer contribute to the de-velopment of the high-saturation hydrate ring and areas of hydratedissociation migrate dynamically, radially from the borehole overtime (see Table 2). This indicates that the secondary hydrates maydissociate again with ongoing heat transfer. Water and gas fromthe dissociation of original and secondary hydrate were squeezeddeeper into the surrounding sediment by the mud invasion and thesecondary hydrates reformed again (Fig. 5c). With weakening ofthe mud invasion and heat transfer, the maximum saturation degreeof the hydrate ring at 24 h fell below that at 12 h.

In summary, hydrate dissociation and reformation mainly oc-curred within 0.25 m of the borehole. Thereinto, hydrate dissocia-tion mainly took place within 0.1 m and no hydrates existed within0.05 m of borehole when mud invasion took place (after 24 h inthis case). The degree of hydrate dissociation around the boreholewas limited because the mud temperature crossed the hydrate sta-bility phase boundary under the conditions of drilling mud pressure(13.54 MPa) and salinity (3.05 per cent). The mud invasion extendedto about 1.0 m from the borehole after 24 h (Fig. 5d). These distancesfrom the borehole (over which dissociation, reformation and mudinvasion occurred) indicate the region in which we should evaluatethe effect of mud invasion on borehole stability and well logging inthe GMGS-1 hydrate drilling project.

3.2 Effect of mud invasion on borehole stability in theGMGS-1 Project

The above simulation results indicate that mud invasion can causethe hydrate dissociation and further change the properties of GHBS,especially the distributions of hydrate saturation, pore pressure andpore-water salinity. As mentioned above, hydrate dissociation canresult in a marked reduction in the mechanical strength of sedi-ments (Winters et al. 1999a,b, 2007; Waite et al. 2008). In ad-dition, mud invasion and hydrate dissociation result in increasedwater content and pore pressure around the borehole. The increasedpore pressure is similar to ‘overpressure’ or ‘excess pore pressure’(Xu & Germanovich 2006; Kwon et al. 2008; Rutqvist & Moridis

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1375

Figure 5. Properties of sediment during mud invasion as a function of distance from the borehole. (a) Distribution of pore pressure around the borehole. (b)Distribution of temperature around the borehole. (c) Distribution of free gas saturation, aqueous saturation and hydrate saturation around the borehole. Thesimulation results for the first 24 h show that the three-phase saturation only changed within 1 m of the borehole. To clearly show variations in the saturationof the three phases, the horizontal axis was set within the range of 1 m. The vertical dashed line shows a region of coexisting hydrate–aqueous–gas. A hydrate‘crest’, induced by hydrate reformation, was even higher than the original hydrate saturation of 0.44 (see the bottom frame). The crest also corresponds to awater trough in the middle frame. (d) Distribution of NaCl concentration around the borehole. After 12 h, the degree of hydrate reformation was enhanced,resulting in a decrease in water content and a corresponding increase in NaCl concentration.

2009; Kwon et al. 2010), defined as the gas pressure above theinitial water pressure prior to dissociation (Holtzman & Juanes2011). The increasing pore pressure results in reduced effectivestress and may destabilize the sediment (Xu & Germanovich 2006;Waite et al. 2008) according to the theory of poromechanics (Biot1941), resulting in instability of the borehole. Especially for soft,fine-grained, mud-dominated GHBS, such as at some sites on theIndian continental margin (Collett et al. 2008) and in the SouthChina Sea (Zhang et al. 2007), gas invasion driven by overpres-sure (caused in turn by mud invasion and hydrate dissociation)may even lead to fracturing in the GHBS (Holtzman & Juanes2010, 2011). Therefore, mud invasion coupled with hydrate disso-

ciation poses a greater risk of borehole stability than in conventionaloil/gas wells.

According to the simulation results in borehole SH7, hydrate dis-sociation mainly took place, and pore pressure and water saturationshowed marked increases within 0.1 m of the borehole (Figs 5a andc). Consequently, this was the most unstable region in the area ofthe borehole. The next unstable region was that at the outside of thehydrate reformation area, that is, the second area of hydrate dissoci-ation. The practical drilling results in the borehole SH7 of GMGS-1 project also showed that gas was released during the drillingoperations. In addition, during open-hole logging with teleme-try, orientation and a four-arm calliper string in GMGS1-SH7,

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1376 F. Ning et al.

Table 2. Hydrate saturation near the borehole at 12, 20 and 24 h. Theoriginal hydrate saturation was 0.44. The data in grey cells are higher thanoriginal hydrate saturation. It can be found that hydrate dissociation takeplace both sides of the high saturation hydrate ring at the same time.

Time (h)

ShPosition r

(m) 0.257240.23040.203870.177640.151710.12606

0.438570.438570.438600.438680.461690.0822112

0.438680.438720.438810.44960.37095020

0.438810.438880.441260.46120.28632024

Figure 6. Schematic diagram of hydrate behaviour, along a radial directionfrom the borehole, during mud invasion in the case that the mud temperatureis higher than the hydrate equilibrium temperature. (a) Initial stage of mudinvasion, showing minimal dissociation of hydrates. (b) With ongoing mudinvasion and heat transfer, a large amount of hydrate is dissociated and theresulting gas is displaced along the pore throat, finally accumulating at theposition where flow resistance is greatest. (c) Hydrate reformation occursin the gas accumulation region, effectively dividing the area of hydrate dis-sociation into two parts. The hydrate reformation reduces the permeabilityand increase the flow resistance of sediment, further promoting the accumu-lation of more gas and water which leads to the high-saturation hydrate ringformation in this area.

spurious caliper 2 and 4 readings were observed in the GHBS.Upon return to the surface, it was found that caliper arms 2 and4 were no longer attached to the calliper sub. The caliper-loggingcurves indicated that borehole shrinkage occurred from the depth

Figure 7. Caliper, sonic and dual-laterolog logging curves of a gas-hydrate-bearing layer in hole SH 7 (Nakai et al. 2007). CAL is four-arm caliperlogging. DLL is duel-laterolog logging. The caliper-logging curves indicatecollapse and shrinkage from the depth of 162.5 mbsf. Consequently, theborehole could not be logged below this depth. Below this depth, the valuesof shallow and deep lateral resistivity began to diverge, with the formerbeing higher than the latter. We concluded that the hydrates began to greatlydecompose around the borehole from this depth. Consequently, the boreholecould not be logged below 170 mbsf.

of 162.5 mbsf (Fig. 7). At the simulation depth (166 mbsf), thevalue of borehole shrinkage is just approximately equal to the mostunstable region which is within 0.1 m of borehole. Tension readingson the logging cable also indicated borehole collapse in the GHBSduring logging with telemetry, full wave sonic and duel-densitystring. The borehole could not be logged below 170 mbsf because ofthese borehole stability issues (Nakai et al. 2007). While for SH1,SH6 and SH9 near by SH7 (Fig. 3), no gas hydrate occurrenceswere observed or indicated by any analyses such as core IR image,porewater freshening of cores and wireline logs. There have nocaliper anomalies and wireline logging can be performed withoutborehole stability issues (Fig. 8). Therefore, we speculate that theborehole instability of Site SH7 was likely induced by mud invasioncoupled with hydrate dissociation, which weaken the strength ofclay-rich sediments and cause the creep shrinkage of clay layersbecause of water absorbing.

3.3 Effect of mud invasion on resistivity logging in theGMGS-1 Project

Another considerable effect of mud invasion is on well logging in theGHBS. Similarly to ice, hydrate is taken to have a non-conductingcomposition in the conduction model for resistivity well logging.The sediment resistivity increases when hydrates are present, whichresults in the high apparent resistivity in term of logging responsecharacteristics (Winters et al. 2007; Gabitto & Barrufet 2009; Lee& Collett 2009a). When the mud salinity is equal to the pore-watersalinity of in situ GHBS (e.g. the case of Site SH7 in GMGS-1project) or lower than the salinity, a higher resistivity girdle aroundthe borehole (compared with the original sediment resistivity) mayform because of the formation of a high-saturation hydrate ring, thepresence of dissociated free gas and the dilution of water salinitydue to mud invasion and hydrate dissociation. The girdle woulddistort the well-logging data for example resistivity logging andprobably result in a higher value of resistivity compared with theoriginal level. Subsequently, a deviation is induced into the calcu-lated hydrate saturation based on Archie’s formula (Archie 1942;we assume that this formula, used for conventional oil/gas reser-voirs, is suitable for unconventional gas hydrate reservoirs. Most

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1377

Figure 8. Caliper, sonic and dual-laterolog logging curves of a gas-hydrate-bearing layer in hole SH 1, SH 6 and SH 9 (Nakai et al. 2007). (a) SH 1, (b) SH 6,(c) SH 9.

natural gas hydrate deposits are still evaluated using this formula)and the Indonesian equation (Poupon & Leveaux 1971), as follows:

Archie′sformula : SH = 1 −(

a Rw

φm Rt

)1/n

, (2)

Indonesianequation :1√Rt

=(

V 1−Vsh/2sh√

Rsh

− φm/2

√a Rw

)Sn/2

w , SH

= 1 − Sw, (3)

where Rw is the resistivity of connate water, a and m are Archieconstants and � is the porosity. The parameter m is commonlycalled the cementation factor and both Archie constants a and mcan be derived empirically (Lee & Collett 2009a). The parametern is derived empirically and Rt is the formation resistivity in thepresence of gas hydrate or other hydrocarbons (Archie 1942). Vsh

is the shale content and Rsh is the shale resistivity. According to eqs(2) and (3), an increase in Rt will cause a corresponding decreasein Sw; consequently, the calculated hydrate saturation is higher thanthat of actual hydrate-bearing sediments.

In practical logging operations in SH7, a dual-laterolog tool wasused to measure both deep and shallow lateral resistivity plus spon-taneous potential. The depth of investigation for deep and shallowlateral resistivity was 0.91 and 0.19 m, respectively (Nakai et al.2007). The present simulation results show that the main area ofdistortion caused by mud invasion coupled with hydrate dissocia-tion and reformation was within 1 m of the borehole (Figs 5c and d).Hydrate dissociation and reformation occurred mainly within 0.25m of borehole (Fig. 5c). The higher resistivity girdle also likelydeveloped in this region. Therefore, shallow lateral resistivity maybe strongly distorted by the mud invasion, whereas deep lateralresistivity is only weakly affected.

To further clarify this point, eq. (2) can be rewritten as follows:

Rt = a Rw

(1 − Sh)nφm(4)

The difference value (�Rt) between shallow lateral resistivity(Rtshal) and deep lateral resistivity (Rtdeep) in GHBS is

�Rt = Rtshal − Rtdeep = a

φm

[Rwshal

(1 − Sshal)n− Rwdeep

(1 − Sdeep)n

], (5)

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1378 F. Ning et al.

Table 3. Calculation parameters for eq. (5), obtained from simulation results at 24 h (Fig. 5). The hydrate saturationof 0.4485 at a shallow logging depth is higher than the original value of 0.44, while the hydrate saturation of 0.4396 ata deep logging depth is close to the original value. The former implies hydrate reformation and the development of ahigh-saturation hydrate ring, while the later implies very weak hydrate dissociation at this depth.

Logging position (m) Pressure (MPa) Temperature (◦C) NaCl concentration (per cent) Hydrate saturation

0.19 (shallow) 13.5068 14.0846 2.9250 0.44850.91 (deep) 13.3512 13.7346 3.0497 0.4396

where Rwshal and Rwdeep are the resistivities of pore water at a shal-low investigation depth of 0.19 m and a deep investigation depth of0.91 m, respectively. Sshal and Sdeep are the corresponding hydratesaturations. Rwshal and Rwdeep can be calculated by using Fofonoff’sequation (Fofonoff 1985). The calculation parameters in eq. (5),listed in Table 3, are obtained from the simulation results for bore-hole SH7 at 24 h (Fig. 5). The hydrate values in Table 3 show thatthere exist a strong hydrate reformation behaviour at shallow depthand weak hydrate dissociation at deep depth. In addition, values ofa = 0.90, m = 1.12 (because it is unconsolidated clay-rich hydrate-bearing sediment in SH7 site) and n = 1.9386 are used for eq. (5)in this paper. The calculated values of Rwshal and Rwdeep are 2.21and 2.06 � m, respectively, and �Rt is 0.15 � m. This result isconsistent with logging data, which show that the shallow investi-gation resistivity is about 0.14 � m larger than that at deep levels(i.e. at the simulation depth; Fig. 7; Nakai et al. 2007). However, forthe Site SH1, SH6 and SH9 without hydrate occurrence, the deepresistivity is almost equal to or even larger than shallow one (Fig. 8).Hence, combining the caliper-logging curve and the dual-laterologlogging curves, we conclude that the behaviour of drilling mud in-vasion is linked to hydrate dissociation and reformation in GHBSof GMGS1-SH7. It would better adopt the deep resistivity log datato evaluate the hydrate reservoir in Site SH7.

Nevertheless, weak hydrate dissociation occurs at the depth ofdeep resistivity logging (Table 3), the calculated hydrate saturationat this depth by using the equation of state of sea water by Fofonoff(1985) and Archie’s formula is still a little higher than that withoutconsidering mud invasion and hydrate dissociation. This differenceis also validated by the pore-water freshening test. The hydrate sat-uration determined by the test was commonly lower than those fromresistivity logging. The maximum value of hydrate saturation frompore-water freshening test is 43 per cent while the peaking valuefrom resistivity logging is about 48 per cent (Fig. 9; Nakai et al.2007; Lu et al. 2008). Lee & Collett (2011) concluded that valuesof gas hydrate saturation derived from resistivity would be overes-timated for anisotropic GHBS (e.g. containing vertical fractures) ifisotropic GHBS were wrongly assumed. However, gas hydrates atSite SH7 are disseminated throughout the sediment, in which nofractures occur. Therefore, the overestimation of hydrate saturationin the GMGS-1 project was most likely due to drilling mud invasioncoupled with hydrate dissociation.

The same phenomenon was observed in the Indian National GasHydrate Program Expedition 01 (NGHP-01): gas hydrate satura-tions (≥50 per cent) calculated by conventional Archie analysisusing logging data were much higher than those (≤26 per cent)estimated from pressure cores from nearby wells in the Krishna–Godavari Basin (Lee & Collett 2009a). In addition to the anisotropicnature of the reservoir and the values of selected parameters (suchas m and n in Archie’s formula), it is also possible that the dis-tortion caused by mud invasion coupled with hydrate dissociationcontributed to the saturation difference. In JIP Leg I, values of gashydrate saturation derived from seismic data were systematicallylower than those estimated from downhole logs (Dai et al. 2008).

In addition to thin layer effect on hydrate saturation estimation(Dai et al. 2008), mud invasion may also have contributed to thisdiscrepancy. Therefore, appropriate correction for the distortion ofwell-logging results by mud invasion should be considered in furtherresearch, in order to correctly identify and assess hydrate resources.At the same time, the drilling conditions should be identify andlimited to reduce the risk of mud invasion and hydrate dissociationduring drilling operations in GHBS, especially in Shenhu hydratearea of China South Sea.

4 D I S C U S S I O N

During mud invasion, secondary hydrates tend to form in areasof hydrate dissociation. The secondary hydrates, together with theoriginal hydrate deposits, may result in the total hydrate satura-tion exceeding that of the in situ GHBS. Therefore, compared withconventional oil/gas-bearing sediments, hydrate dissociation andreformation are the main characteristics of mud invasion in GHBSwhen the invasion conditions are in the unstable region of gas hy-drates. The characteristics of drilling mud invasion in conventionaloil and gas reservoirs are influenced by many factors. For example,invasion depth and range are controlled by the properties of themud and the reservoir. In order to identify the influence factors ofmud invasion in GHBS and limit the suitable drilling conditionsfor borehole stability and good logging interpretation, we furtherdiscuss the effects of drilling mud properties on mud invasion inGHBS, based on Site SH7 of the GMGS-1 hydrate drilling project.

4.1 Effect of mud density on mud invasion in GHBS

In some cases, mud density needs to be increased for drilling safety,such as to prevent wellbore instability and well kick. A typical ex-ample is well WR313-G, which was drilled during the GOM JIP LegII. In this case, it was found to be difficult to use sea water with inter-mittent sweeps of 10.5 lbm gal–1 (about 1258 kg m–3) drilling mudto maintain borehole stability and to remove drill cuttings from thehole below 503 mbsf. Multiple backreams per stand had to be usedto prevent the drill string from occasional pack-off. After adoptinga protocol of continuously pumping 10.5 lbm gal–1 of weighteddrilling fluid, the rest of the hole was drilled incident free (Collettet al. 2012). This protocol actually increased the equivalent circu-lation density of drilling mud; therefore, higher mud density resultsin enhanced borehole stability. However, it may also influence thebehaviour of mud invasion in GHBS.

Under the same mud temperature (15◦C) and salinity(3.05 per cent) as above, the behaviours of mud invasion into GHBSwere simulated under different pressure scenarios. The scenariosconsidered are densities of drilling fluid of 1190, 1250 and 2000kg m–3, equal to pressures of 13.54, 13.64 and 15 MPa at the in-vasion spot, respectively. The simulation results revealed that nogas hydrate dissociation or reformation took place for a density of2000 kg m–3, because hydrates remained stable under this condi-tion, for which the equilibrium temperature is about 15.7◦C at apressure of 15 MPa, and the NaCl concentration is 3.05 per cent.

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1379

Figure 9. GMGS1-SH7 gas-hydrate saturation from pore-water freshening. The hydrate saturation was below the value of 48 per cent obtained from resistivitylogging (Nakai et al. 2007).

No free gas or regenerated hydrates formed around the borehole(Fig. 10a).

However, if the mud temperature remains higher than the hy-drate stability condition under higher pressures (e.g. 1250 kg m–3 at15◦C or even 2000 kg m–3 at 20◦C), the mud invasion shows someinteresting characteristics. Higher pressures led to deeper invasion(see the NaCl concentration plot in Fig. 10b). If the difference be-tween mud temperature and hydrate equilibrium temperature is verysmall, there is a low degree of gas hydrate dissociation. The higherpressure acted to retard the hydrate dissociation and led to a smalleramount of free gas content in the formation (see the gas saturationplot in Fig. 10b), which caused a lower degree of hydrate reforma-tion. The maximum hydrate saturation in the case of 1190 kg m–3 at15◦C was about 0.4496, and the value at the same temperature for1250 kg m–3 was about 0.4468 (see the hydrate saturation plot ofFig. 10b). However, in the case of a larger difference between themud temperature and the hydrate equilibrium temperature, higherpressure led to deeper invasion, increased the speed and depth ofheat transfer and extended the range of hydrate dissociation (seeFig. 10b for the case of 2000 kg m–3 at 20◦C).

Of relevance to the simulation results is the fact that gas is com-pressible and its occurrence results in an increase in multiphaseflow resistance. For the case of 2000 kg m–3 at 20◦C, the largeramount of free gas that dissociated from hydrates might be moreeasily accumulated and detained in certain small regions of thehydrate dissociation area during mud invasion, compared with thecase of a smaller amount of gas (Fig. 6). In addition, the higher

pressure resulted in pressure propagation over a greater distance.These factors led to the rapid formation of secondary hydrateswith higher saturation and greater range. The ‘hydrate crest’ hada saturation of about 0.634 at 0.27 m from the borehole. The po-sition of hydrate crest corresponded to an ‘aqueous and gas sat-uration trough’ and a ‘NaCl concentration crest’. The exothermicreaction of hydrate reformation would also result in enhanced hy-drate dissociation at sites behind the area of hydrate reformation,resulting in the formation of a deeper ‘hydrate trough’ (see thediscussion in Section 3.1 and Fig. 6) and a corresponding ‘aque-ous and gas saturation crest’ and a ‘NaCl concentration trough’(Fig. 10b).

The formation of secondary hydrate with high saturation alsohad a strong influence on the mud invasion in GHBS. Whenthe density of mud increased to 2000 kg m–3 at 20◦C, the porepressure at the hydrate crest (about 0.27 m from the borehole)showed a sharp decrease (Fig. 10c), indicating that the gas contentshowed a marked decrease and that fluid flow was blocked by high-saturation hydrate. The results of both experiments and numericalsimulations indicate that the permeability of sediments is reducedby hydrate formation (Liu & Flemings 2007; Garg et al. 2008;Kneafsey et al. 2011).

The above results show that an increase in mud density inhibitshydrate dissociation and enhances borehole stability under condi-tions of a low degree of hydrate dissociation. Therefore, the muddensity to prevent borehole instability in Site SH7 should be largerthan 1190 kg m–3, perhaps 1250 kg m–3 is suitable, as its results in

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1380 F. Ning et al.

Figure 10. Properties of sediment around the borehole for various mud densities. (a) Distribution of gas saturation around the borehole at same temperatureand 24 h. (b) Distributions of hydrate saturation, aqueous saturation, gas saturation and NaCl concentration around the borehole at 20 h. (c) Distribution ofpore pressure around the borehole at 20 h. The red curve indicates that the pore pressure spread was retarded by high-saturation hydrate, which reduced thepermeability of GHBS.

borehole WR313-G of JIP II project. However, if the mud temper-ature greatly exceeds the hydrate equilibrium temperature, a highermud density would enhance the degree of hydrate dissociation,reformation and mud invasion, which would cause problems forborehole stability and well logging interpretation. Furthermore, anexcessively high mud density (e.g. 2000 kg m–3) would give rise toother drilling problems, such as mud loss, given the narrow win-dow of safe mud density in offshore drilling operations. Oceanichydrates commonly occur in shallow and underconsolidated sedi-ments, resulting in a high risk of formation fracture. Hence, it isnoteworthy that the mud density of 2000 kg m–3 discussed here israrely adopted in oceanic hydrate drilling operations. However, thepresent case shows that control of drilling mud temperature is animportant issue for safe drilling in gas hydrate and for accurate well

logging. Next, we discuss the effect of mud temperature on mudinvasion into GHBS.

4.2 Effect of mud temperature on mud invasion in GHBS

During conventional oil and gas drilling, the invasion of drilling mudinto sediment is governed by differences in hydraulic and osmoticpressure, and the effect of temperature on mud invasion is neglected(Zhang et al. 1999), meaning that the mud invasion is treated as anisothermal process (Bilardo et al. 1996). This isothermal case isinvalid when the mud invasion takes place under the condition ofmud temperature higher than the hydrate equilibrium temperaturein the drilling of GHBS. Figs 11a and b show that the higher thetemperature of drilling mud, the larger the degree of hydrate disso-

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1381

Figure 11. Properties of sediment around the borehole at 24 h for various mud temperatures. (a) Distributions of hydrate, aqueous and gas saturation aroundthe borehole. (b) Distributions of temperature, pore pressure and NaCl concentration around the borehole.

ciation and the greater extent of mud invasion for a given density(1190 kg m–3) and salinity (3.05 per cent). Moreover, the larger thetemperatures difference between mud and sediment, the greater theheat transfer into sediment during mud invasion. It is more difficultto form a high-saturation secondary hydrate ring by using highertemperature mud (20◦C in Fig. 11a). In addition, the greater the hy-drate dissociation, the more water is released, and the stronger thediluting effect on the salinity of pore water (Fig. 11b). Thereby thecirculating mud should be kept cool and a fast cycle speed shouldbe adopted. However, low-temperature mud circulation may resultin the gas released from hydrate dissociation to reform as hydratein the borehole. To address this problem, kinetic inhibitors or anti-agglomerants could be added to the mud instead of thermodynamicinhibitors. The reason why thermodynamic inhibitors are not beingconsidered will be discussed in the following section.

4.3 Effect of mud salinity on mud invasion in GHBS

In deep-water oil or gas drilling, hydrate inhibitors (e.g. NaCl andKCl) are usually added to the mud to enhance its performanceand to prevent gas hydrate formation and aggregation in the well(Ebeltoft et al. 1997). Here we simulated the effect of mud salinityon mud invasion into GHBS. Higher mud salinity results in greaterhydrate dissociation during mud invasion under a given mud den-sity (1190 kg m–3) and temperature (15◦C; Fig. 12a). This behaviourcan be explained as follows: (1) salt, as a thermodynamic inhibitor,

can cause the phase equilibrium curve to shift toward destabiliza-tion, which enhances dissociation and (2) increasing salinity resultsin enhanced heat transfer because of the higher heat conductivityof salts compared with surrounding sedimentary rocks (Nagiharaet al. 2002). A high-saturation hydrate ring and corresponding low-saturation water ring form around the borehole with the invasion ofhigh-salinity mud (Fig. 12a). There is greater hydrate dissociationand more secondary hydrates, in the case of mud with a salinity of10 per cent compared with a salinity of 6 per cent. The temperatureof sediments near the wellbore showed a marked drop due to theendothermic reaction associated with hydrate dissociation (Masudaet al. 1999, Nazridoust & Ahmadi 2007) and due to the higherthermal conductivity of mud with higher salinity (Fig. 12b). Thedropped temperature keeps the hydrates which occur outside of thehigh-saturation hydrate ring stable. Therefore, no hydrate troughwas observed in Fig. 12a. The temperature increased again due tothe exothermic reaction of secondary hydrate formation (Fig. 12b).For a mud salinity of 10 per cent, pressure propagation was delayedby secondary hydrate formation during mud invasion (Fig. 12b),thereby blocking further mud invasion. The higher the mud salinity,the more pronounced this trend.

If mud salinity is much higher than the salinity of the originalsediment, mud invasion would still increase the sediment salinityand appear as a ‘low-resistivity invasion’ even though the waterfrom hydrate dissociation would dilute the salinity of pore water(Fig. 12b). However, the high-saturation hydrate ring and gas from

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1382 F. Ning et al.

Figure 12. Properties of sediment around the borehole after 5 h for various mud salinities. (a) Distributions of hydrate, aqueous and gas saturation around theborehole. (b) Distributions of temperature, pore pressure and NaCl concentration around the borehole.

hydrate dissociation would act to increase the sediment resistivity.Therefore, the dynamic response characteristics of resistivity welllogging, in the case of mud invasion in the GHBS, are more compli-cated than those in conventional sediment. Appropriate correctionfor the distortion of well-logging results by mud invasion should beconsidered in further research, in order to correctly identify and as-sess hydrate resources. Nevertheless, a general principle is to adoptlower (like sea water) or no salts in the drilling mud for the hydratedrilling.

5 C O N C LU S I O N S A N D S U G G E S T I O N SF O R F U T U R E W O R K

In this study, the characteristics of drilling mud invasion and the ef-fect of mud properties on mud invasion into GHBS was investigatedby numerical simulations under the conditions of mud temperatureover hydrate equilibrium temperature and overbalanced drilling,considering the case of the GMGS-1 project in China. The primaryresults are as follows:

(1) The radial formation of secondary hydrate at a certain depthin sediments adjacent to a borehole is attributed to the increasedpore pressure caused by mud invasion and flow resistance, to theendothermic cooling that accompanies hydrate dissociation com-pounded by the Joule–Thompson effect, and to the lagged effectof heat transfer in sediments. The hydrate saturation of secondary

hydrates and existing in situ hydrates could be higher than that ofthe original sediment, resulting in the formation of a high-saturationhydrate ring around the borehole.

(2) In the case of the borehole SH7 of GMGS-1 project, thesimulation results, practical drilling and logging data indicated theinfluence of mud invasion coupled with hydrate dissociation andreformation on well logging and borehole stability was confinedto within 1 m of the borehole. The likely occurrence of a high-saturation hydrate ring, the presence of dissociated free gas and thedilution of water salinity associated with hydrate dissociation couldaffect the well-logging data and cause an erroneously high valueof shallow laterolog compared with greater depths. Subsequently,the hydrate saturation calculated from resistivity logging would behigher than that of actual hydrate-bearing sediments. Therefore,suitable correction for the distortion of well-logging results by mudinvasion should be considered when identifying and refining oceanhydrate resources.

(3) Mud density, temperature and salinity are the important fac-tors affecting hydrate stability and secondary hydrate formationduring mud invasion, among which temperature and salinity playprimarily roles. It is necessary to control the temperature, density,salinity and filtration loss of mud and to prevent hydrate dissociationto ensure borehole stability, drilling safety and well-logging accu-racy when drilling in GHBS. Managed pressure drilling (MPD)operation can be employed to ensure a small pressure difference

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

Mud invasion in gas-hydrate-bearing sediments 1383

between mud and sediments. In addition, it would be preferentialto adopt a low-salinity mud system containing suitable temporaryplugging additives for drilling. The simulation results also showthat, under some conditions, even the LWD method may not beapplicable to GHBS because mud invasion and hydrate dissocia-tion occur extremely rapidly. This problem may be addressed bydeveloping a novel deep lateral logging method.

A C K N OW L E D G E M E N T S

The authors would like to thank Dr. George Moridis and WenyueXu for valuable suggestions regarding our models and theoreticalanalysis. We also sincerely thank Dr. Timothy S. Collett and M.W.Lee for kind and valuable discussions at the 7th ICGH. We alsothank two anonymous reviewers for their constructive suggestionsand Mrs. Valerie Dennis and Dr. Bruce Buffett as editors. Thiswork was supported by the National Natural Science Foundation ofChina (Nos 50704028, 40974071, U0933004), the Guangzhou Cen-ter for Gas Hydrate Research, Chinese Academy of Sciences (No.o807s2) and Fundamental Research Funds for Central Universities(No. CUGL100410 & 120112).

R E F E R E N C E S

Archie, G., 1942. The electrical resistivity log as an aid in determining somereservoir characteristics, Petrol. Transact., 146, 54–62.

Bilardo, U., Alimonti, C., Chiarabelli, A. & Caetani, F.C., 1996. Formationwater saturation from drilling fluid filtrate invasion: comparison of dis-placement modelling and induction well log response, J. Petrol. Sci. Eng.,15, 251–259.

Biot, M.A., 1941. General theory of three-dimensional consolidation, J.appl. Phys, 12, 155–164.

Birchwood, R., Singh, R. & Mese, A., 2008. Estimating the in situ mechan-ical properties of sediments containing gas hydrates, in Paper presentedat the 6th International Conference on Gas Hydrates (ICGH 2008), Van-couver, British Columbia, Canada.

Birchwood, R. & Noeth, S., 2012. Horizontal stress contrast in the shallowmarine sediments of the Gulf of Mexico sites Walker Ridge 313 andAtwater Valley 13 and 14—geological observations, effects on wellborestability, and implications for drilling, Mar. Petrol. Geol., 34(1), 186–208.

Booth, J.S., Rowe, M.M., Fischer, K.M. & Survey, G., 1996. Offshore gashydrate sample database: with an overview and preliminary analysis, U.S.Dept. of the Interior, U.S. Geological Survey.

Chand, S., Minshull, T.A., Priest, J.A., Best, A.I., Clayton, C.R.I. & Waite,W.F., 2006. An effective medium inversion algorithm for gas hydrate quan-tification and its application to laboratory and borehole measurements ofgas hydrate-bearing sediments, Geophys. J. Int., 166, 543–552.

Chi, S., Torres-Verdin, C., Wu, J.H. & Alpak, F.O., 2006. Assessment of mud-filtrate-invasion effects on borehole acoustic logs and radial profiling offormation elastic properties, SPE Reserv. Eval. Eng., 9, 553–564.

Collett, T.S., 2001. A review of well-log analysis techniques used to assessgas-hydrate-bearing reservoirs, Geophys. Monogr., 124, 189–210.

Collett, T.S. & Lee, M.W., 2000. Reservoir characterization of marine andpermafrost associated gas hydrate accumulations with downhole welllogs, Ann. N. Y. Acad. Sci., 912, 51–64.

Collett, T.S., Riedel, M., Cochran, J.R., Boswell, R., Kumar, P. & Sathe, A.the NGHP Expedition 01Scientific Party, 2008. Indian continental margingas hydrate prospects: results of the Indian National Gas Hydrate Pro-gram (NGHP) expedition 01, in Paper presented at the 6th InternationalConference on Gas Hydrates (ICGH 2008), Vancouver, British Columbia,Canada.

Collett, T.S., Boswell, R., Mrozewski, S., Guerin, G., Cook, A., Frye,M., Shedd, W. & McConnell, D., 2009. Gulf of Mexico Gas Hy-drate Joint Industry Project Leg II—operational summary, in Pro-ceedings of the Drilling and Scientific Results of the 2009 Gulfof Mexico Gas Hydrate Joint Industry Project Leg II. Available

at http://www.netl.doe.gov/technologies/oil-gas/publications/Hydrates/2009Reports/OpSum.pdf.

Collett, T.S., Lee, M.W., Zyrianova, M.V., Mrozewski, S.A., Guerin, G.,Cook, A.E. & Goldberg, D.S., 2012. Gulf of Mexico Gas Hydrate JointIndustry Project Leg II logging-while-drilling data acquisition and anal-ysis. Mar. Petrol. Geol., 34, 41–61.

Dai, J.C., Banik, N., Gillespie, D. & Dutta, N., 2008. Exploration for gas hy-drates in the deepwater, northern Gulf of Mexico: part II. Model validationby drilling, Mar. Petrol. Geol., 25, 845–859.

Dickens, G.R. & QuinbyHunt, M.S., 1997. Methane hydrate stability inpore water: a simple theoretical approach for geophysical applications, J.geophys. Res.-Solid Earth, 102, 773–783.

Ebeltoft, H., Yousif, M. & Soergard, E., 1997. Hydrate control during deep-water drilling: overview and new drilling-fluids formulations, in Paperpresented at the SPE Annual Technical Conference and Exhibition, SanAntonio, Texas.

Fofonoff, N.P., 1985. Physical properties of seawater: a new salinity scaleand equation of state for seawater, J. geophys. Res., 90(C2), e3332–e3342.

Freij-Ayoub, R., Tan, C., Clennell, B., Tohidi, B. & Yang, J.H., 2007. Awellbore stability model for hydrate bearing sediments, J. Petrol. Sci.Eng., 57, 209–220.

Gabitto, J. & Barrufet, M., 2009. Gas Hydrates Research Programs: AnInternational Review Prairie View, A&M University, United States.

Garg, S.K., Pritchett, J.W., Katoh, A., Baba, K. & Fujii, T., 2008. Amathematical model for the formation and dissociation of methane hy-drates in the marine environment, J. geophys. Res.-Solid Earth, 113,doi:10.1029/2006JB004768.

Holtzman, R. & Juanes, R., 2010. Crossover from fingering to frac-turing in deformable disordered media, Phys. Rev. E, 82, 046305,doi:10.1103/PhysRevE.82.046305.

Holtzman, R. & Juanes, R., 2011. Thermodynamic and hydrodynamic con-straints on overpressure caused by hydrate dissociation: a pore-scalemodel, Geophys. Res. Lett., 38, L14308, doi:10.1029/2011GL047937.

Hu, H., Tang, H., Luo, J. & Wei, H., 2009. Deepwater gas hydrates drillingand coring techniques, Oil Drill. Prod. Technol., 31, 27–30.

Jaiswal, P., Zelt, C.A. & Pecher, I.A., 2006. Seismic characterization of a gashydrate system in the Gulf of Mexico using wide-aperture data, Geophys.J. Int., 165, 108–120.

Jiang, G.S., Liu, T.L., Ning, F.L., Tu, Y.Z., Zhang, L., Yu, Y.B. & Kuang,L.X., 2011. Polyethylene glycol drilling fluid for drilling in marine gashydrates-bearing sediments: an experimental study, Energies, 4, 140–150.

Johnson, A., Patil, S. & Dandekar, A., 2011. Experimental investigation ofgas-water relative permeability for gas-hydrate-bearing sediments fromthe Mount Elbert Gas Hydrate Stratigraphic Test Well, Alaska NorthSlope, Mar. Petrol. Geol., 28, 419–426.

Kneafsey, T.J., Seol, Y., Gupta, A. & Tomutsa, L., 2011. Permeability oflaboratory-formed methane-hydrate-bearing sand: measurements and ob-servations using x-ray computed tomography, SPE J., 16, 78–94.

Kneafsey, T.J., Tomutsa, L., Moridis, G.J., Seol, Y., Freifeld, B.M., Taylor,C.E. & Gupta, A., 2007. Methane hydrate formation and dissociationin a partially saturated core-scale sand sample, J. Petrol. Sci. Eng., 56,108–126.

Konno, Y., Masuda, Y., Takenaka, T., Oyama, H., Ouchi, H. & Kurihara, M.,2008. Numerical study on permeability hysteresis during hydrate disso-ciation in hot water injection, in Paper presented at the 6th InternationalConference on Gas Hydrates(ICGH 2008), Vancouver, British Columbia,Canada.

Kvenvolden, K.A., 1993. Gas hydrates-geological perspective and globalchange, Rev. Geophys., 31, 173–187.

Kwon, T.H., Cho, G.C. & Santamarina, J.C., 2008. Gas hydrate dissocia-tion in sediments: pressure-temperature evolution, Geochem. Geophys.Geosyst., 9, Q03019, doi:10.1029/2007GC001920.

Kwon, T.H., Song, K.I. & Cho, G.C., 2010. Destabilization of marine gashydrate-bearing sediments induced by a hot wellbore: a numerical ap-proach, Energy Fuel, 24, 5493–5507.

Lee, M.W. & Collett, T.S., 2008. Integrated analysis of well logs and seismicdata to estimate gas hydrate concentrations at Keathley Canyon, Gulf ofMexico, Mar. Petrol. Geol., 25, 924–931.

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

1384 F. Ning et al.

Lee, M.W. & Collett, T.S., 2009a. Gas hydrate saturations estimated fromfractured reservoir at Site NGHP-01–10, Krishna-Godavari Basin, India,J. geophys. Res.-Solid Earth, 114, B07102, doi:10.1029/2008JB006237.

Lee, M.W. & Collett, T.S., 2009b. Unique problems associated with seismicanalysis of partially gas-saturated unconsolidated sediments, Mar. Petrol.Geol., 26, 775–781.

Lee, M.W. & Collett, T.S., 2011. Three types of gas hydrate reservoirs inthe gulf of mexico identified in LWD data, in Paper presented at the7th International Conference on Gas Hydrates (ICGH 2011), Edinburgh,Scotland, United Kingdom.

Lee, M.W., Collett, T.S. & Lewis, K.A., 2012. Anisotropic models to accountfor large borehole washouts to estimate gas hydrate saturations in the Gulfof Mexico Gas Hydrate Joint Industry Project Leg II Alaminos Canyon21 B well, J. Mar. Petrol. Geol., 34, 85–95.

Li, G., Moridis, G.J., Zhang, K. & Li, X.S., 2010. Evaluation of gas produc-tion potential from marine gas hydrate deposits in Shenhu area of SouthChina Sea, Energy Fuels, 24, 6018–6033.

Liu, X.L. & Flemings, P.B., 2007. Dynamic multiphase flow model of hy-drate formation in marine sediments, J. geophys. Res.-Solid Earth, 112,B03101, doi:10.1029/2005JB004227.

Lu, J.A., Yang, S.X., Wu, N.Y., Zhang, G.X., Zhang, M. & Liang, J.Q., 2008.Well logging evaluation of gas hydrates in shenhu area, south china sea,Geoscience, 22, 447–451.

Masuda, Y., Fujinaga, Y., Naganawa, S., Fujita, K., Sato, K. & Hayashi,Y., 1999. Modeling and experimental studies on dissociation of methanegas hydrates in Berea sandstone cores, in Paper presented at the 3rdInternational Conference on Gas Hydrates(ICGH 1999), Salt Lake City,Utah.

Moridis, G.J., Kowalsky, M.B. & Pruess, K., 2007. Depressurization-inducedgas production from class 1 hydrate deposits, SPE Reserv. Eval. Eng., 10,458–481.

Moridis, G.J., Kowalsky, M.B. & Pruess, K., 2009a. TOUGH+HYDRATEv1.1 User’s Manual: A Code for the Simulation of System Behavior inHydrate-Bearing Geologic Media, Lawrence Berkeley National Labora-tory, University of California, Berkeley, California.

Moridis, G.J., Reagan, M.I., Kim, S.J., Seol, Y. & Zhang, K., 2009b. Eval-uation of the gas production potential of marine hydrate deposits in theUlleung Basin of the Korean East Sea, SPE J., 14, 759–781.

Moridis, G.J. & Reagan, M.T., 2007. Gas production from oceanic class 2hydrate accumulations, in Paper presented at the Offshore TechnologyConference, Houston, Texas.

Moridis, G.J., Seol, Y. & Kneafsey, T.J., 2005. Studies of reaction kinetics ofmethane hydrate dissocation in porous media, in Paper presented at the5th International Conference on Gas Hydrate(ICGH 2005), Trondheim,Norway.

Nagihara, S., Brooks, J.M., Bernard, B.B., Summer, N., Cole, G. & Lewis,T., 2002. Application of marine heat flow data important in oil, gas ex-ploration, Oil Gas J., 100, 43–49.

Nakai, T.C., Tjok, K.M. & Humphrey, G., 2007. Deepwater gas hydrateinvestigation Shenhu survey area south South China Sea, offshore China.Factual Field Report. Report No.0201–6130. Guangzhou Marine Geolog-ical Survey, Guangzhou, P.R.China.

Nazridoust, K. & Ahmadi, G., 2007. Computational modeling of methanehydrate dissociation in a sandstone core, Chem. Eng. Sci., 62, 6155–6177.

Ning, F.L., Jiang, G.S., Zhang, L., Wu, X., Dou, B. & Tu, Y.Z., 2008a. Anal-ysis of key factors affecting wellbore stability in gas hydrate formations,Petrol. Drill. Tech., 36, 59–61.

Ning, F.L., Jiang, G.S., Zhang, L., Dou, B. & Wu, X., 2008b. Analysison characteristics of drilling fluids invading into gas hydrates-bearingformation, in Paper presented at the 6th International Conference on GasHydrates (ICGH 2008), Vancouver, British Columbia, Canada.

Ning, F.L., Wu, N.Y., Jiang, G.S. & Zhang, L., 2009. The effect of gas hy-drates dissociation and drilling fluids invasion upon borehole stability inoceanic gas hydrates-bearing sediment, in Paper presented at the Ameri-can Geophysical Union, Fall Meeting 2009, San Francisco, California.

Ning, F.L., Wu, N.Y., Jiang, G.S., Zhang, L., Guan, J.A., Yu, Y.B. & Tang,F.L., 2010. A method to use solar energy for the production of gas frommarine hydrate-bearing sediments: a case study on the Shenhu area, En-ergies, 3, 1861–1879.

Phelps, G.D., 1995. Computation of mud filtrate invasion profiles, J. Can.Petrol. Technol., 34, 18–27.

Poupon, A. & Leveaux, J., 1971. Evaluation of water saturation in shalyformations, Log. Analyst., 7, 1–8.

Priest, J.A., Best, A.I. & Clayton, C.R.I., 2006. Attenuation of seismic wavesin methane gas hydrate-bearing sand, Geophys. J. Int., 164, 149–159.

Riedel, M., Long, P.E. & Collett, T.S., 2006. Estimates of in situ gas hy-drate concentration from resistivity monitoring of gas hydrate bearingsediments during temperature equilibration, Mar. Geol., 227, 215–225.

Rutqvist, J., Grover, T. & Moridis, G.J., 2008. Coupled hydrologic, ther-mal and geomechanical analysis of well bore stability in hydrate-bearingsediments, in Paper presented at the Offshore Technology Conference,Houston, Texas.

Rutqvist, J. & Moridis, G.J., 2009. Numerical studies on the geomechanicalstability of hydrate-bearing sediments, SPE J., 14, 267–282.

Rutqvist, J., Moridis, G.J., Grover, T. & Collett, T., 2009. Geomechanicalresponse of permafrost-associated hydrate deposits to depressurization-induced gas production, J. Petrol. Sci. Eng., 67, 1–12.

Seol, Y. & Myshakin, E., 2011. Experimental and numerical observationsof hydrate reformation during depressurization in a core-scale reactor,Energy Fuel, 25, 1099–1110.

Sloan, E.D., 2003. Clathrate hydrate measurements: microscopic, meso-scopic, and macroscopic, J. Chem. Thermodyn., 35, 41–53.

Spangenberg, E., 2001. Modeling of the influence of gas hydrate contenton the electrical properties of porous sediments, J. geophys. Res.-SolidEarth, 106, 6535–6548.

Stone, H., 1970. Probability model for estimating three-phase relative per-meability, J. Petrol. Technol., 22, 214–218.

van Genuchten, M.T., 1980. A closed-form equation for predicting the hy-draulic conductivity of unsaturated soils, Soil Sci. Soc. Am. J, 44, 892–898.

Waite, W.F., Kneafsey, T.J., Winters, W.J. & Mason, D.H., 2008. Physicalproperty changes in hydrate-bearing sediment due to depressurization andsubsequent repressurization, J. geophys. Res.-Solid Earth, 113, B07102.

Waite, W.F. et al., 2009. Physical properties of hydrate-bearing sediments,Rev. Geophys., 47, 1–38.

Weitemeyer, K.A., Constable, S. & Trehu, A.M., 2011. A marine electro-magnetic survey to detect gas hydrate at Hydrate Ridge, Oregon, Geophys.J. Int., 187, 45–62.

Winters, W.J. et al., 1999a. Physical properties of sediments from theJAPEX/JNOC/GSC Mallik 2L-38 gas hydrate research well, Bull.-Geol.Surv. Can., 544, 95–100.

Winters, W.J., Pecher, I.A., Booth, J.S., Mason, D.H., Relle, M.K. & Dillon,W.P., 1999b. Properties of samples containing natural gas hydrate from theJAPEX/JNOC/GSC Mallik 2L-38 gas hydrate research well, determinedusing Gas Hydrate and Test Laboratory Instrument (GHASTLI), Bull.-Geol. Surv. Can., 544, 241–250.

Winters, W.J., Waite, W.F., Mason, D.H., Gilbert, L.Y. & Pecher, I.A., 2007.Methane gas hydrate effect on sediment acoustic and strength properties,J. Petrol. Sci. Eng., 56, 127–135.

Wu, N.Y., Yang, S.X., Wang, H.B., Liang, J.Q., Gong, Y.H., Lu, Z.Q.,Wu, D.D. & Guan, H.X., 2009. Gas-bearing fluid influx sub-system forgas hydrate geological system in Shenhu Area, J. Geophys., 52, 1641–1650.

Xu, W.Y. & Germanovich, L.N., 2006. Excess pore pressure resulting frommethane hydrate dissociation in marine sediments: a theoretical approach,J. geophys. Res.-Solid Earth, 111, B01104, doi:10.1029/2004JB003600.

Zhang, H.Q. et al., 2007. Successful and surprising results for China’sfirst gas hydrate drilling expedition: Fire in the ice. Methane HydrateNewsletter, National Energy Technology Laboratory, US Department ofEnergy, Fall issue 6–9.

Zhang, J.H., Hu, Q. & Liu, Z.H., 1999. Estimation of true formation resis-tivity and water saturation with a time-lapse induction logging method,Log. Analyst., 40, 138–148.

Zhang, K., Moridis, G.J., Wu, Y.S. & Pruess, K., 2008. A domain de-composition approach for large-scale simulations of flow processes inhydrate-bearing geologic media, in Paper presented at the 6th Interna-tional Conference on Gas Hydrates (ICGH 2008), Vancouver, BritishColumbia, Canada.

at University of C

alifornia, Berkeley on O

ctober 13, 2014http://gji.oxfordjournals.org/

Dow

nloaded from