Salt caverns: the key to storing hydrogen?

Pro­jec­tions show that by 2050 the ener­gy con­sumed in Europe will be divid­ed equal­ly between elec­tric­i­ty and hydro­gen – replac­ing fos­sil fuels that cur­rent­ly account for the largest share. It is there­fore crit­i­cal for us to start get­ting a hold of how to man­age this gas on a large scale, now. In par­tic­u­lar, the entire chain hydro­gen ener­gy (pro­duc­tion, trans­port, dis­tri­b­u­tion and end-use) will require sig­nif­i­cant stor­age capac­i­ty, par­tic­u­lar­ly if the pri­ma­ry pro­duc­tion meth­ods are inter­mit­tent, as is the case for wind tur­bines and solar panels.

To do this, we will need a sys­tem capa­ble of con­nect­ing the pro­duc­tion equip­ment, such as elec­trol­y­sers, with the dis­tri­b­u­tion net­work. This sys­tem will need to be able to store hydro­gen en masse when it is pro­duced and con­se­quent­ly make it avail­able when the grid requires. To achieve this, we pro­pose to store H₂ in deep saline cav­i­ties, and with the hydro­gen indus­try gain­ing momen­tum, we esti­mate that by 2050 Europe could need sev­er­al hun­dred of them.

Salt cav­erns are wide­ly used in Europe today, par­tic­u­lar­ly for stor­ing methane (‘nat­ur­al gas’). They take advan­tage of the pres­ence of salt lay­ers or domes in the sub­soil sev­er­al hun­dred metres thick that extend over large areas. In France, the total exten­sion of salt cav­erns is cur­rent­ly the order of 20,000 km². The salt, which is not very per­me­able, can be eas­i­ly dis­solved to cre­ate cav­erns with oil wells to dig the rock down to the salt for­ma­tion. Then, after some time, a cav­ern with a typ­i­cal size of 500,000m³ is avail­able. In the­o­ry, we could sim­ply use these pre-exist­ing cav­erns to store hydro­gen. In the­o­ry, they would be able to stor­ing around 6,000 tonnes of hydro­gen at vary­ing pres­sures of 6–24MPa – oper­at­ed like an oxy­gen tank for diving.

Today in France there are about thir­ty gas stor­age cav­i­ties, spread over three sites. Some have been in oper­a­tion for about fifty years. There are sev­er­al hun­dred in the world, includ­ing a few hydro­gen stor­age cav­erns already used by the chem­i­cal indus­try. Sev­er­al indus­tri­al pilot projects for the pro­duc­tion and use of hydro­gen are organ­ised around some of these exist­ing salt cav­erns. In France, there is the Hyp­ster project led by Storengy in Etrez and sup­port­ed by the Euro­pean Union, Hygéo by Tere­ga, HdF and BRGM in Car­resse-Cass­aber and Hygreen led in Manosque by Storengy-Geo­stock. If these projects exist, it is because there are still some chal­lenges to over­come if we are to achieve large-scale hydro­gen stor­age in salt caverns.

Even for con­ven­tion­al struc­tures, social accept­abil­i­ty is the major chal­lenge posed by the con­struc­tion of new large-scale ener­gy facil­i­ties (nuclear pow­er plants, dams, wind farms). As such, it is up to the design­ers to iden­ti­fy and explain to the pub­lic the par­tic­u­lar­i­ties of the struc­tures from the point of view of their safe­ty and the solu­tions pro­vid­ed. Salt cav­erns are no excep­tion. Since the hydro­gen mol­e­cule is par­tic­u­lar­ly mobile, the major prob­lem is the seal­ing of the met­al access shaft, which is sev­er­al kilo­me­tres long. A num­ber of acci­dents or inci­dents from the past are known and well described in the 2,000 or so salt cav­erns for the stor­age of liq­uid or gaseous hydro­car­bons in oper­a­tion around the world. For the most part, these inci­dents occurred a long time ago, as the indus­try pro­gres­sive­ly adopt­ed a prin­ci­ple known as the “dou­ble bar­ri­er” where­by the pres­sure is con­tin­u­ous­ly mon­i­tored to detect when the gas has breached a first bar­ri­er; there­by avoid­ing the risk of breach­ing the second.

Anoth­er essen­tial check is the “leak test”. This con­sists of low­er­ing a col­umn of nitro­gen a lit­tle above the roof of the cav­ern and mon­i­tor­ing the evo­lu­tion of the gas-brine inter­face: a rapid rise is a sign of poor seal­ing. On the scale of the cave, the well is a very thin cap­il­lary, and the sys­tem resem­bles an extreme­ly sen­si­tive barom­e­ter or ther­mome­ter. The chal­lenge is to track down small leaks, of the order of 10^-4/year of the stored vol­ume. An abun­dant lit­er­a­ture, to which the Sol­id Mechan­ics Lab­o­ra­to­ry (LMS) has made a major con­tri­bu­tion, is devot­ed to this test and we can expect, with hydro­gen stor­age, impor­tant devel­op­ments con­cern­ing the method, peri­od­ic­i­ty, and accept­abil­i­ty criteria.

Over large time scales, salt is a vis­cous liq­uid and there­fore any cav­i­ty will grad­u­al­ly close. The result­ing annu­al loss of vol­ume must remain below one per­cent to pre­vent the stor­able vol­ume from decreas­ing too quick­ly, espe­cial­ly for the deep­est cav­i­ties. But also because it can cause dam­age to the wall or the salt-access shaft inter­face. The descrip­tion of the behav­iour of salt is old but has been pro­found­ly renewed; Rock Physics estab­lish­es that there is a spe­cif­ic mech­a­nism of defor­ma­tion under very low stress­es, which is dif­fi­cult to mea­sure because the defor­ma­tion speeds involved are of the order of 10^-12/s. How­ev­er, salt is also a brit­tle mate­r­i­al, which is sus­cep­ti­ble to break­age – par­tic­u­lar­ly under the effect of a sud­den change in mechan­i­cal load. Sud­den vari­a­tions in stock and there­fore in pres­sure can be expect­ed if the cav­i­ties are fed by high­ly inter­mit­tent hydro­gen pro­duc­tion and must sat­is­fy a demand that is also discontinuous.

The uses of hydro­gen for mobil­i­ty requires extreme puri­ty. How­ev­er, the gas will have to remain for a long time in a cav­i­ty at the bot­tom of which there are thou­sands of m³ of brine. The brine con­tains sul­phates from the anhy­drite (H₂S) fre­quent­ly asso­ci­at­ed with under­ground salt. The gas is wet and loaded with var­i­ous impu­ri­ties includ­ing H₂S, which is par­tic­u­lar­ly harm­ful for down­stream gas uses. Purifi­ca­tion can be a major expense.

Final­ly, giv­en its large size, the cav­ern is a com­plex ther­mo­dy­nam­ic machine – one can speak of cave mete­o­rol­o­gy with rain, snow, and tem­per­a­ture inver­sion, made very exot­ic by the very large changes in pres­sure. The dri­ving force behind this machine is the pres­ence of a nat­ur­al tem­per­a­ture gra­di­ent in the sub­soil; it gen­er­ates intense con­vec­tion of gas in the upper part of the cave, coun­tered in the low­er part by the main­te­nance of a rel­a­tive­ly cold tem­per­a­ture of the brine result­ing from the suc­ces­sion of episodes of vapor­i­sa­tion and con­den­sa­tion of the water. These phe­nom­e­na, high­light­ed by the LMS with Storengy, have con­se­quences for the puri­ty of the gas extract­ed that have yet to be ful­ly explored.

Interview by James Bowers