Creative Writing

A natural hydrate dissolution experiment on complex multi-component hydrates on the sea floor

Description
Dissolution of natural hydrate cores was measured using time-lapse photography on the seafloor at Barkley Canyon (850 m depth and 4.17 °C). Two types of hydrate fabrics in close contact with one another were studied: a "yellow" hydrate
Published
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A natural hydrate dissolution experiment on complexmulti-component hydrates on the sea floor K.C. Hester a,b , E.T. Peltzer a , P.M. Walz a , R.M. Dunk a , E.D. Sloan b , P.G. Brewer a,* a Monterey Bay Aquarium Research Institute, 7700 Sandholdt Road, Moss Landing, CA 95039, USA b Center for Hydrate Research, Colorado School of Mines, 1500 Illinois Street, Golden, CO 80401, USA Received 21 January 2009; accepted in revised form 11 August 2009; available online 18 August 2009 Abstract Dissolution of natural hydrate cores was measured using time-lapse photography on the seafloor at Barkley Canyon(850 m depth and 4.17   C). Two types of hydrate fabrics in close contact with one another were studied: a  “ yellow ”  hydratestained with condensate oil and a  “ white ”  hydrate. From thermogenic srcins, both fabrics contained methane as well as hea-vier hydrocarbons. These multi-component hydrates were calculated to be well within  p  –  T   stability conditions (<200 m waterdepth needed at 4.17   C). While stable in pressure and temperature, the hydrates were bathed in under-saturated seawater,which promoted dissolution. The flux of gas from the shrinking yellow hydrate core was 0.15 ± 0.01 mmol gas/m 2 s, whilethe white hydrate dissolved faster at 0.25 ± 0.02 mmol gas/m 2 s. To determine the controlling mechanism for the observedchanges in the hydrate cores, experimental results were compared with an engineering correlation for convective mass trans-fer. Using water velocity as a fitting parameter, the correlation agreed well with results from a previous dissolution experimenton well-characterized synthetic hydrates. Even with a number of other unknowns, when applied to the natural hydrate, themass transfer correlation predicted the dissolution rate within 20%. This seafloor-based experiment, along with visual obser-vations of seafloor hydrate dissolution over a 3-day period, were used to further understand the fate of natural seafloorhydrates exposed on the seafloor. By showing that mass transfer is the rate-controlling mechanism for dissolution of thesenatural hydrate outcrops, proper hydrodynamic calculations can be employed to give a refined estimate on hydrate dissolu-tion rates.   2009 Elsevier Ltd. All rights reserved. 1. INTRODUCTION Gas hydrates are naturally occurring compounds foundin the World’s permafrost and oceans (Sloan and Koh,2008). In the natural environment, sufficient temperatureand pressure conditions exist for hydrate formation overa vast portion of the ocean. However, in addition to pres-sure and temperature, the chemical potential of the gas inthe hydrate must be equal to the surrounding waters. If the gas concentration in surrounding water is under-satu-rated with respect to the hydrate guest, the hydrate will dis-solve due to the chemical disequilibrium (Rehder et al.,2004).Naturalhydratesareacurrenttopicofresearchbecauseof theirpossibleuseasafutureenergyresourceandtheirpoten-tialimpactsonclimatechangeandseafloorstability.Naturalhydratedepositstypicallyexistinsedimentsbelowthesulfatereduction zone and canpersistto the base of the hydrate sta-bility zone (Buffett, 2000). These hydrates are thought to beinsulated from small variations in pressure and temperature(Xu et al., 2001; Reagan and Moridis, 2007). However, thereare other oceanic hydrates which exist very close to the sea-floor in areas of high fluid flux (Tre´hu et al., 2006). In somecases, hydrate mounds were found exposed on the seafloor(Sassen et al., 2001; Chapman et al., 2004). Unlike deeperhydrate deposits, near-seafloor hydrates can be subjected tosignificant variations in temperature, water velocities, gas 0016-7037/$ - see front matter    2009 Elsevier Ltd. All rights reserved.doi:10.1016/j.gca.2009.08.007 * Corresponding author. Tel.: +1 831 775 1706; fax: +1 831 7751620. E-mail address:  brpe@mbari.org (P.G. Brewer). www.elsevier.com/locate/gca  Available online at www.sciencedirect.com Geochimica et Cosmochimica Acta 73 (2009) 6747–6756  flux, and changing benthic communities inhabiting andencrusting the site (Macdonald et al., 1994).Gas hydrates in contact with under-saturated ambientseawater will dissolve over time. This is of particular rele-vance to hydrate mounds on the seafloor; thus if the hy-drate accumulation is to be sustained, they must becontinuously supplied with a supporting flux from below.However, to date, little work has been done on the rate atwhich this dissolution occurs. The majority of hydrate dis-sociation work has focused on heat-transfer controlled sys-tems, more relevant to oil and gas pipeline blockages(Davies et al., 2006). The only field study to report a mea-sured dissolution rate was performed on laboratory-synthe-sized sI CH 4  and CO 2  hydrates exposed to ambient under-saturated seawater around 1000 m depth in the MontereyBay (Rehder et al., 2004). It was concluded from theseexperiments that mass transfer was controlling the dissolu-tion process, as the rate of dissolution was proportional tothe individual guest molecules water solubility. However,the experimental rate of dissolution for these synthetic sam-ples appeared to be too rapid compared with some in situobservations of hydrate mounds. For example, little visualchange was observed over a 350-day period for a hydratemound at Bush Hill, Gulf of Mexico (Vardaro et al.,2006). Similar hydrate mounds in the Bush Hill areashowed significant changes over a 10-month period (Mac-donald et al., 1994). Open questions remain as to thedynamics of these seafloor hydrate outcroppings. In orderto understand this situation better, a controlled field exper-iment was performed on natural hydrates at Barkley Can-yon. The results of which may then be compared to anychanges observed of the natural exposures over time so asto gain understanding of the balance of geochemical fluxesrequired to create and maintain these systems. 2. GEOLOGICAL SETTING Barkley Canyon (Fig. 1) is located off Vancouver Islandon the northern Cascadia Margin accretionary prism. Theoccurrence of hydrates there was first discovered by a fish-ing trawler, and the hydrates were soon shown to exist asmounds on the seafloor (Spence et al., 2001). At around850 m depth, the areal extent of the hydrate field is esti-mated to be 0.5 km 2 with hydrate both exposed and veiledwith a thin layer of sediment, on the order of centimeters inthickness (Chapman et al., 2004). A light yellow condensatefluid is present in the surrounding sediment and associatedwith the hydrate, causing yellow staining (Lu et al., 2007).With the buoyant hydrate faces exposed on the seafloorreaching 7 m in length and 3 m in height and with only athin sediment cover, the hydrate mass must continue deeperto be anchored below the surface veneer of sediments. Pre-vious studies at the site have shown the hydrate methane tohave thermogenic srcins, with both sII and sH hydratestructures present (Lu et al., 2007). 3. METHODS All field measurements were carried out during a surveyof Barkley Canyon conducted on August 9–17, 2006,aboard the MBARI R/V Western Flyer using the remotelyoperated vehicle (ROV) Tiburon. Exposed outcrops of gashydrates were cored using a specially constructed stainlesssteel coring device, operated by the vehicle robotic arm,with an internal radius of 2.1 cm. The leading edge of thecorer had a serrated rim. Inside the corer was a hydraulicram to eject the hydrate core.Hydrate samples were cored directly using the ROVmanipulator arm and then injected into a sampling cell.Measurements were performed to determine hydrate com-position and dissolution rate. It should be noted that differ-ent cores necessarily were used for the characterization of the hydrate fabrics (composition and porosity) and dissolu-tion. While the heterogeneous nature of these hydrates isknown, for our analysis, we must assume that the resultsare transferable between cores sampled within close prox-imity of each other (within 5–10 cm as shown in Fig. EA-1).In order to determine the gas compositions of the twofabrics, a gas funnel containing a 240VDC 400 W heatingelement was used. A hydrate core was injected into the in-verted heated funnel. The heater then caused hydrate to dis-sociate allowing the gas to collect in the top of the funnel.The gas was collected inside 150 cm 3 evacuated stainlesssteel containers. Laboratory gas chromatography (GC)was later performed on both gases collected from the yellowand white hydrate fabrics (Peltzer et al., 2006).For the hydrate dissolution experiment, the hydrate corewas expelled into an open mesh exposure container, whichallowed for exposure to ambient benthic currents while pre-venting the hydrate from floating away. No rotation of thecores was observed in the time-lapse photographs over thecourse of the experiment.In order to observe the slow dissolution of the hydrate inseawater at Barkley Canyon, we elected to use time-lapsephotography. The hydrates were measured for a total timeof around 54 h. This technique was similar to the approachfor measuring the slow dissolution of synthetic hydratespecimens in Monterey Bay (Rehder et al., 2004). Thetime-lapse photography was performed using a NikonCool-pix 3 megapixel camera set to record an image at10-min intervals. The camera, along with lighting and a bat-tery pack, was mounted on a custom aluminum frame. Theopen mesh exposure chamber was positioned on a pre-determined mount for optimal illumination and resolution.In addition to the time-lapse photographs, frame grabsfrom the high definition (HD) video recorded on multiplevisits by the ROV were also analyzed. The frame was lo-cated in close proximity to the hydrate mounds (<25 m)and near the seafloor (<2 m). The ambient conditions were4.17   C, 868 dbar (8.8 MPa), and  S   = 34.3.Changes in the hydrate core over the dissolution processwere determined in post-cruise analysis of the time-lapsephotographs and HD frame grabs. Two independent anal-yses were performed using Photoshop  and ImageJ(Abramoff et al., 2004) to determine the hydrate coredimensions as a function of pixel number. The pixel lengthscale was calibrated in each photograph based on a knowndimension of the open mesh container. Each length anddiameter of a hydrate core was taken as an average of threemeasurements across the core. The measurement uncer- 6748 K.C. Hester et al./Geochimica et Cosmochimica Acta 73 (2009) 6747–6756  tainty for both the volume and diameters of the hydratecores was taken as the standard deviation of these threemeasurements. Propagation of errors was used to estimateuncertainties in other calculated properties, such as dissolu-tion rate.The molar gas flux into the surrounding ocean based onmeasured dissolution rate was calculated by  DR ¼ SR 2   q hyd  n þ 1  ð 1 Þ where  DR  is the dissolution rate in mol gas/(m 2 s),  SR  is themeasured shrinking rate of the cores in mm/s,  q hyd   is themolar hydrate density, and  n  is the overall hydration num-ber. The density and overall hydration number were calcu-lated based on the measured hydrate composition given inthe next section. All calculations of gas and hydrate prop-erties were performed using MultiFlash from Infochem. 4. RESULTS TheBarkleyCanyonareawasmarkedwithnumeroushy-drate outcroppings. Hydrates in these outcroppings wereeither directly exposed tothe surrounding seawater orveiledwithathinlayerofsediment.Avisualsurveyofthesemoundsshowed characteristics reported for other seafloor hydrateoccurrences, such as in the Gulf of Mexico. Along with thepresence of chemosynthetic communities (vesicomyid clamsand bacterial mat), several exposed hydrate outcroppingswereobservedtobeundercutwithirregularcrevasses.Thesefeatures support the theory that these outcroppings wereformed beneath the seafloor and are being pushed upwardsovertime(Sassenetal.,1994;Vardaroetal.,2006).Thevent-ingofgaswasnotobservedduringthisexpedition.However,gasandoilbubbleswereobservedrisinginthewatercolumnwhen the ROV disturbed the surrounding sediment. Fig. 1. Map of Barkley Canyon located off Vancouver Island. Topology from a MBARI AUV mapping survey in August 2006 (Caress et al.,2006).Natural hydrate dissolution experiment 6749  One of the hydrate mounds, dubbed  “ Cliff Hanger ” , wasselected as the primary experimental site. Two distinct hy-drate fabrics were present in the hydrate mound as shownin Fig. 2: a  “ yellow ”  hydrate, stained with a light conden-sate, and a  “ white ”  hydrate, which was not associated withthe condensate phase. Similar to other hydrate mounds, theyellow hydrate was undercut with the white hydrate fillingthe space below. Comparing the two hydrate fabrics, theyellow hydrate was much more resistant to coring. Whena fractured sample of a yellow hydrate core was placed ina glass cell, the condensate was observed to float to thetop of the cell; the separation of the condensate and hydraterevealed a white/clear hydrate. This indicated that the con-densate was trapped in veins throughout the hydratemounds, not as bubbles or pockets in the bulk hydrate.The white hydrate was cored relatively quickly and smallgas bubbles were observed escaping during the coring. 4.1. Hydrate fabric and gas composition For hydrate dissolution in under-saturated seawater, thediffusional driving force is determined using the hydratecomposition. As given in Table 1, while very similar in bulkcomposition, the yellow hydrate was slightly more enrichedin heavy hydrocarbons, such as ethane, propane, and bu-tanes. From the gas compositions and ambient seafloorconditions, properties of the hydrate were calculated withproperties of the hydrate given in Table 2.In addition to obtaining the gas composition, the coredissociation in the gas funnel allowed for volume estimatesof the gas evolved. Both white (Fig. 3a) and yellow (Fig. 3c) hydrate cores were inserted into the gas funnel. The initialimages allowed for estimates of the overall core volume.Following complete dissociation of the hydrate using theheater, the evolved gas was contained in the top of the fun-nel for both the white (Fig. 3b) and yellow (Fig. 3d) hydrate cores. As the hydrate neared complete visual dissociation,the heater was turned off. Because hydrate dissociation isan endothermic process and the gas funnel was bathed inambient seawater for over 10 min following dissociation,we assumed the gas to be at the ambient temperature of 4.2   C. The headspace gas volume was later estimated bymatching the images with measured volumes of water inthe laboratory. We assumed all gas evolved from the hy-drate or in the pore space would be present in the gas headspace. While some gas will be absorbed by the surroundingwater, we assume this is negligible based on low gas solubil-ity (less than 1 g/kg H 2 O) and the short time frame of theexperiment (around 10 min).Because both the volume of the core and resulting gashead space were measured, it is possible to estimate theporosity of the hydrate. For the white hydrate core, thegas head space was 125 ± 5 cm 3 . From the calculatedin situ gas density of 5.06 mol/L (using the measured hy-drate gas composition at seafloor conditions), the gas headspace contained 0.63 ± 0.03 mol of gas. With a bulk corevolume of 103.5 ± 3.3 cm 3 , the calculated gas evolution if  Fig. 2. The  “ Cliff Hanger ”  site at Barkley Canyon. This mound had two distinct hydrate fabrics present: a yellow oil-stained fabric overlayinga white more-porous fabric. Separate hydrate cores were taken from both faces of the two fabrics for the composition and dissolutionmeasurements. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)Table 1Measured composition of the two hydrate fabrics.Yellow hydrate White hydrateMethane 76.97 86.20Ethane 10.65 7.04Propane 7.87 3.01 i  -Butane 1.10 0.52 n -Butane 0.69 0.61Neopentane 0.13 0.14 i  -Pentane 0.09 0.12 n -Pentane 0.08 0.09N 2  + O 2  0.15 0.50CO 2  0.50 0.09Table 2Hydrate properties calculated based on ambient seafloor conditionsand the measured hydrate composition using MultiFlash. Porosityestimated based on method described in Section 4.1.Yellow hydrate White hydrateHydrate density (mol/L) 50.87 51.07 n  6.34 6.25Porosity (%) 10 13–476750 K.C. Hester et al./Geochimica et Cosmochimica Acta 73 (2009) 6747–6756  the core was pure solid hydrate (51.07 mol/L,  n  = 6.25)would have been 0.73 ± 0.02 mol of gas. As no condensatewas present in the white hydrate, the most likely explana-tion for this discrepancy is porosity, with the pore spacesfilled with either water or gas. This is in agreement withthe small gas bubbles observed while coring the whitehydrate.To give an estimate of porosity, the bulk core volumewas considered to be the sum of hydrate and porosity (filledwith water and/or free gas). The minimum porosity, if thepores were only filled with water, would be around 13%.This is because the water in the pores does not contributeto the head gas volume. The maximum porosity, if thepores were only filled with gas, would be around 47%.Based on previous work, it is likely a combination of gasand water in the pores resulting in porosity between 13%and 47%.For the yellow hydrate core, the total volume of the hy-drate core was determined to be 111.9 ± 3.8 cm 3 . After dis-sociation, the gas head space was 120 ± 5 cm 3 . At in situconditions, the gas density was calculated based on themeasured composition to be 5.94 mol/L, corresponding to0.71 ± 0.06 mol of gas evolved for the hydrate in the exper-iment. In addition to gas, a layer of condensate oil was pres-ent at the gas/liquid interface when the yellow hydrate wasdissociated. Based on the thickness of this layer, theamount of condensate in the core was 11.4 cm 3 . The yellowhydrate has a calculated hydrate density of 50.87 mol/L and Fig. 3. Images from gas funnel experiments to collect hydrate gas. The cores of both white (a) and yellow (c) hydrate are initially inserted intothe gas funnel. A heating element then dissociated the core, with the gas headspace filled with gas liberated from the white (b) and yellow (d)hydrate cores. A condensate layer was also present following dissociation of the yellow core. (For interpretation of the references to color inthis figure legend, the reader is referred to the web version of this article.)Natural hydrate dissolution experiment 6751
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks