sign in sign up
<<
Home , Boulby mine, Carnallite, Glauberite, Kieserite, Langbeinite, Leonite

www.saltworkconsultants.com

Salty MattersJohn Warren - Tuesday July 31, 2018 : Geology of an alternate low- fertiliser Introduction with early-formed or (Hardie 1984). Mg and Cl reach high concentrations during the latest Polyhalite is the hydrated of , and stages of evaporation resulting in the precipitation , with the formula (K2Ca2Mg(SO4)4·2H2O). of and (Figure 1). Polyhalite crystallises in the triclinic system, but individual euhedral crystals are very rare in nature where the usual Utilisation habit is fibrous to massive. It is typically colourless, grey Today polyhalite is mined as a prime ore target in only one to white, although the natural colour in some place in the world, the Israel Chemical Limited (ICL)- deposits tends to brick red due to iron oxide inclusions. owned Mine, located below the North Sea, off Primary (syndepositional) precipitates can be layered at a the North coast of the UK. Currently, ICL-UK mm to cm scale. Its Mohs hardness is 3.5 with a specific distributes around 500 kt/y from its as a di- gravity of 2.8. It was first described from a Salzburg mine rect application fertiliser or bulk-blend/compound NPK in 1818 and the name comes from a Latin root that refers additive and is in the process of expanding its polysulphate to the “many ”evident in its chemical formula. output as it downgrades its MOP operations. A second Polyhalite is relatively easy to distinguish from associated major polyhalite mine, located near the Boulby Mine, is by simple field tests. Its hardness separates it from proposed by Sirius PLC. This renewed interest most evaporite salts, other than anhydrite. It has a bitter in polyhalite as an economical source of potash fertiliser is taste and is water soluble (incongruent dissolution), with why I am writing this article. remnants of gypsum and (K2Ca(SO4)2·2H2O), Sulphate of Potash (SOP) fertiliser was initially derived which is also soluble, leaving behind a final residue of gyp- from polyhalite/ in the US during the first half sum (unlike and which dissolve wholly and of the twentieth century. But then during the 1940s, after congruently in fresh water). Polyhalite is not deliquescent, its discovery in vast quantities by prospectors searching for unlike carnallite, and gives a purple flame result when held oil in Saskatchewan, the focus for the world’s potash fer- in a gas flame due to its potassium content, unlike the non-potassium salts. Table 1 lists the constituents of the Anhydrite CaS0 various evaporite salts mentioned in this article. 4 Bischofite MgCl2.6H2O The complete equilibrium evaporation of modern seawa- Bloedite Na2Mg(S04)2 • 4H20 ter at 25ºC produces the sequence: CaCO3 (cal- MgSO4.7H2O cite and aragonite), CaSO4 (gypsum and anhydrite), halite CaSO .Na SO (NaCl), other sulphates (glauberite, polyhalite, epsomite, 4 2 4 Gypsum CaS0 • 2H 0 hexahydrite, and ) and (carnallite and 4 2 bischofite). Halite is the dominant mineral because of the Halite NaCl Hexahydrite MgSO .6H O high concentration of Na and Cl in seawater (Figure 1; 4 2 KMg(S0 )Cl • 3H 0 after Harvie et al., 1980). 4 2 Kieserite MgS0 This mineral sequence with abundant sulphate bitterns re- 4 Langbeinite K Mg (S0 ) flects the relatively high concentration of sulphate in mod- 2 2 4 3 K2Mg(S04)2 • 4H20 ern seawater (molar SO4 > Ca), which, following evapora- Schoenite () K2Mg(S04)2 • 6H20 tion and precipitation of CaCO3 and CaSO4, produces a Polyhalite K Ca Mg(S0 ) • 2H 0 SO4-rich, Ca-depleted at halite saturation and be- 2 2 4 4 2 yond, from which Mg-sulphate bitterns precipitate. Poly- Sylvite KCl halite can form syndepositionally at temperatures below 30 Syngentite K2Ca(S04)2 • H20 °C, via back-reaction of the evaporating K-Mg-SO4 brine Table 1. Sulphate and other evaporite minerals mentioned in this article. Page 1 www.saltworkconsultants.com

Gypsum Anhydrite log moles H2O Halite +2 +1 0 Glauberite -1 Polyhalite 0 EP-HX-KS Halite Carnallite Bischo te

+1 halite -1 ppts. bischo te ppts. hexahydrite to kieserite epsomite to hexahydrite Hexahydrite Kieserite epsomite ppts. 0 polyhalite consumed - Carnallite Cl anhydrite to (solution no longer Anhydrite glauberite -2 + invariant) Anhydrite Na

anhydrite polyhalite (cumulative) -1 2+ carnallite log molality log moles ppt’d log moles salt ppt’d Mg ppts. and gypsum to solution -3 anhydrite becomes 2- invariant SO4 Gypsum K+ ppts -2 Ca2+ Final invariant -4 composition Halite Bischo te Epsomite

Gypsum Carnallite Polyhalite Glauberite rst appearance -3 of kieserite 1 2 5 10 20 50 100 200 500 1000 Evaporative concentration factor +2 +1 0 -1 -3

A. B. log moles H2O

Figure 1. Theoretical evaporation series using modern seawater composition including concepts of back reaction during concentration (after Harvie et al., 1980). Note the predominance of sulphate salts in the bittern stage and the onset of polyhalite precipitation in the early bittern brine field. tiliser supply moved to Canada and the mining of sylvite phate®. Sirius plans to market its polyhalite as POLY4®. (muriate of potash- KCl), (Warren, 2016). The geology of CRU estimates the global consumption of potassi- SOP was discussed in an earlier Salty Matters article (May um-magnesium-sulphate (SOPM) fertilisers in 2017 at 15, 2015). All aspects of SOP and MOP geology and min- 1.7 Mt total product; a comparatively small total compared ing are discussed in detail in Chapter 11, Warren (2016). to the widely traded 65.5 Mt (MOP) ICL currently markets crushed and processed polyhalite market (https://www.crugroup.com "Will polyhalite dis- from the Boulby mine as polysulphate (Table 2). The purity rupt the fertiliser industry?” published online April 2018; of the Polysulphate product from the Boulby mine is very last accessed 12 July 2018). Polyhalite accounts for around high (95% polyhalite) with <5% sodium chloride (NaCl) 450-460 kt of current SOPM fertilisers. ICL-UK is cur- and traces of boron (B) and iron (Fe) at 300 and 100 ppm, Product Manufacturer (Brand) % K O % MgO % SO % Cl respectively (Yermiyahu et al., 2017). The declared min- 2 3 imum analysis of polyhalite for S, K, Mg and Ca is 48% Polyhalite ICL-UK (Polysulphate®) 14 6 48 <5 sulphur trioxide (SO3), 14% potassium oxide (K2O), 6% Langbeinite Mosaic (K-Mag®) 22 18 55-66 1-3 magnesium oxide (MgO) and 17% calcium oxide (CaO), Intrepid (Trio®) respectively. This compares to a K2O content of contains Hartsalz K+S 11 5 10 45 60–63% in MOP and around 50% in SOP. (Magnesia-Kainit®) Schoenite Chinese producers 21-24 5-6 35-40 2-3 Polyhalite’s make up in terms of K, Mg, S and Cl pro- Blends portions is similar to the other major potassium-magne- sium-sulphate (SOPM) fertilisers: Langbeinite, Schöenite MOP +Kieserite K+S (Korn-Kal®) 40 6 12 38 and Patentkali® (Table 2). All are marketed as low-chloride SOP + Kieserite K+S (Patenkali®) 30 10 42 >3 potash fertilisers with additional magnesium and sulphur Kalimagnesia® components. ICL-UK markets polyhalite as a multi-nutri- Table 2. Listing of the main manufacturers of SOPM fertilisers currently available, along with K O contents. ent, low-chloride fertiliser under the brand name Polysul- 2 Page 2 www.saltworkconsultants.com

rently ramping-up production to 1 Mt/y (140 kt/y K2O) growth of a potato crop at Tapira Brazil (Figure 2; Mel- by 2020, as it simultaneously phases out MOP production lo et al., 2017). Its impact on tuber starch and tuber dry at the Boulby mine. At current production levels, this will matter exceeded that of either MOP or SOP applications. be equivalent to almost 40% of the current SOPM market Cultivar Asterix at Tapira is mainly used for frying and in K2O terms. chip-making (fries) in the food processing industry. High dry matter and starch content improve texture, and lower Sirius Minerals’ planned Phase I mine capacity (10 Mt/y sugar contents result in less darkening of fries which is de- product) is on a different scale altogether; around four sirable. High dry matter percentage enables lower oil ab- times larger than the current SOPM market in K2O terms. sorption while frying, resulting in lower oil usage per unit This volume is almost the same size as the current global product. Tuber firmness is essential to handle mechanical potassium sulphate (SOP) market, which is the most pop- stresses that may occur during tuber harvesting, transport, ular low-chloride potash fertiliser, outside China. The suc- and storage. Crunchiness and hardness are positively relat- cess of any future expanded SOPM application in agricul- ed to starch and dry matter contents and specific gravity. ture is contentious; the majority of SOPM consumption has traditionally been concentrated close to production It seems polyhalite products are probably suitable as a low sites, nurtured by the local marketing efforts of producers. chloride fertiliser replacement of sylvite in some agricul- The proposed worldwide expansion will be tied to increas- tural applications especially in arid acid, infertile soils as ing acceptance by the agricultural community of polyhalite found in parts of Israel and other dry growing areas in the as an acceptable cheaper substitute for SOP and perhaps Middle East where salinisation due to fertiliser residues is MOP. a known problem. Yermiyahu et al., 2017, found the trans- port and leaching of Ca, Mg, K and S in soil following Polyhalite as a fertiliser. polyhalite application is lower than following the applica- In the last few years, the use of polyhalite as an agricul- tion of the equivalent sulphate salt fertilisers. The residual tural fertiliser has been tested successfully in a number of effect of polyhalite fertiliser on the subsequently grown studies supported by Sirius minerals (Mello et al., 2018; crop is higher than the impact from the equivalent sul- Pavuluri et al., 2017). Polyhalite supplies four nutrients, is phate salts, especially regarding Ca, Mg and S. Irrigation less water soluble than the more conventional potassium management, as determined by the leaching fraction, has a sources and may conceivably provide a slower release of substantial effect on the efficiency of polyhalite as a source nutrients. Studies comparing polyhalite to other K and Mg of K, Ca, Mg and S for plant nutrition. fertilisers have shown that polyhalite is at least as effective Geology of polyhalite as potassium sulphate (K2SO4) as a slow release source of K, and at least as effective as potassium chloride (KCl) plus Polyhalite is a common constituent of many ancient evap- orite sequences, especially in Permian and Neogene de- magnesium sulphate (MgSO4) as a source of K and Mg (Barbarick, 1991). posits, due to evaporation of Na-K-Mg-Cl-SO4 marine . These sulphate bittern assemblages correspond to

The possibility of successful use of polyhalite as a fertiliser periods of MgSO4-enriched ocean chemistries (Lowen- is illustrated by the positive effects of its application on the 16 20 Polyhalite SOP MOP 15 19 14 18 13 12 17 11

Tuber starch % starch Tuber 16 10 SOP 0.01055*x + 10.202 r2=0.83 16 9 dry % matter Tuber 2= Polyhalite 0.01224*x + 10.142 r2=0.65 Polyhalite 0.00768*x + 15.47 r 0.90 8 16 0 62 124 186 248 0 62 124 186 248 K application rate, kg ha-1 K application rate, kg ha-1

Figure 2. Dry matter content and starch in potatoes for three sources over a range of potassium (K) fertilizer application rates at Ta- pira,Brazil. MOP = muriate of potash; SOP = sulfate of potash (after Mello et al., 2017) Page 3 www.saltworkconsultants.com

NE 0 SW normal marine brine and progres- Salt Dyke Sand 7 sively increased in a landward di- pan dunes 1 May 1979 rection, suggesting gypsum disso-

Abandoned 2 lution by groundwater crossflows. 3 inlet 5 6 Isotope study suggests both water 101 4 0 21 and aqueous sulphate in the mud- N 208 43 6 wt 7 5 flat porewater have a mixed marine and continental origin (Figure 4). 500 m A. B. Thus, it appears that sulphate ions 50m are provided in part by marine 10cm brines, in part by continental wa- 5 101 208 ters which have dissolved Pleisto- 0 Halite crust cene interstitial gypsum present at Gypsum depth. The replacement of gypsum wt Nodular polyhalite by polyhalite requires not only 20 wt wt Massive polyhalite high Mg2+ and K+, but also high

Depth in cm Sandy clays SO4 concentrations in the cross- Sands (with marine shells) 40 flowing solu­tions (Braitsch, 1971). C. May 1979 May 1980 April 1981 The polyhalite in Ojo de Lieb- Figure 3. Polyhalite in Ojo de Liebre, Mexico (after Pierre, 1983). A) Schematic map of the salt re mudflats is diagenetic but also pond located at the southeastern margin of the Ojo de Liebre evaporitic complex, showing location­ of the samples collected in May 1979 (0 to 7), in May 1980 (101 to 107) and in April 1981 (201 to penecontemporaneous with the 208). B) Schematic cross-section of the salt pond along its transverse axis, showing the sedimentary crystallisation of gypsum. How- succession in May 1979. C) Evolution of the evaporitic succession in the central part of the salt pond ever, in brines with temperatures during May 1979, May 1980 and April 1981. >30°C, polyhalite may also be a stein et al., 2003; Demicco et al., 2005). primary co-precipitate with ha- lite, as is occurring in recent saltworks near Santa Pola, SE Modern polyhalite occurrences Spain (as observed by B.C. Schreiber), and in cool-zone The presence of syndepositional polyhalite in the supra- (cryogenic) salt lakes associated with widespread mira- tidal evaporite flats around the Ojo de Liebre lagoon was Modern Ojo first discovered by Holser (1966), who attributed its origin 20 seawater to the diagenesis of gypsum by interstitial marine brines. SO4 in pore brines of the A large part of this area is now occupied­ by artificial salt modern sabkha mud at ponds. However, some remnants of the ancient­ evaporite 19 flats are still accessible, for example, on the southeast coast

of the evaporitic complex, where sedimentolog­ical, chem- CD

S 18

ical and isotopic investigations were performed on evapo- 34

ritic sediments and interstitial solutions (Figure 3a, Pierre,  1983). In May 1979, the evaporitic succession was mainly 17 composed of gypsum; a few centimetres below the surface, polyhalite was present in the form of small nodules that May 1979 were partially replacing former gypsum crystals (Figure 16 May 1980 3b,c). In May 1980, this evaporitic succession was drasti- April 1981 Gypsum crystals from cally modified, since polyhalite replaced gypsum sediments last interglacial (5 km inland) lying below the water table. This gives an exact timing for 15 the mineral transformation from gypsum to halite of one 8 9 10 11 12 13 14 18 year to replace 10cm thick interval of gypsum with a 10cm  OSMOW interval of polyhalite, which points to a chemical evolution Figure 4. and sulphur isotope composition of aqueous sulphate of of the solutions permeating the sediments. interstitial solutions collected in the salt pond at the southeastern margin of the Ojo de Liebre evaporitic complex (Pierre, 1983). The oxygen values During this one-year period, ionic concentrations of in- of Sangamonian (last interglacial) gypsum were measured on samples terstitial brine increased from thirteen to 18 times with collected about 6 km south of the margin of the evaporite flats. The in- respect to seawater concentration (Pierre, 1983). SO levels termediary position of the polyhalite samples (shaded brown) between 4 seawater and waters from dissolution of more continental-influenced of interstitial solutions in the sabkha were higher than in older gypsum implies mixing between waters from these two sources.

Page 4 www.saltworkconsultants.com

bilite-glauberite such as in Karabogazgol (Andriyasova, (middle to upper Aragonian). Thick beds of nearly pure 1972). sepiolite were deposited in ponds extended at the toes of arkosic alluviums. Sepiolite is also found within calcrete Polyhalite is also a minor but widespread phase associat- profiles in these environments. Minor amounts of sepio- ed with glauberite in the Late Pleistocene-early Holocene lite are commonly recognised along with palygorskite in sediments of Lop Nur China (Ma et al., 2016). There, the open lacustrine areas. On the other hand, Mg-bentonites natural lake evaporites are nonmarine assemblages of mi- characteristically occur associated with dolostones and rabilite-glauberite-polyhalite-bloedite-gypsum-halite. The fine micaceous sands in sequences that provide evidence evaporitic stages of the lake fill contain massive amounts of of fluctuations in the lake level. Polyhalite typically occurs glauberite and polyhalite compared to the other salts pres- as felty and spherulitic aggregates that alternate with cen- ent. Polyhalite in the upper 40 m of the lake column and timetre-thick halite layers or millimetre-thick glauberite its predominance, is indicative of pervasive back-reactions, laminae in the Lower Saline Unit(Figure 5). The polyhalite as is the presence of very minor amounts of carnallite and sylvite in the same section (Ma et al., 2010; Dong et al., 2012). Ancient occurrences

Most ancient occurrences are interpreted as early diage- Glauberite Magnesite Halite Gypsum Anhydrite Polyhalite Calcite netic, formed in shallow brine crossflows as replacement of anhydrite or gypsum. Even so, there is a direct association between higher volumes of polyhalite in marine evaporite Secondary basins and times of MgSO4 enrichment of ocean waters. gypsum Neogene polyhalite Polyhalite is not found as a widespread primary precipitate in rocks of this age even though ocean chemistries are Mg- Zone Weathering SO4-enriched. Instead, polyhalite is typically a minor but extensive early burial replacement of anhydrite or gypsum. The better-documented examples of this type of replace- Gypsum ment polyhalite are found in association with gypsum and /glauberite in various Tertiary lacustrine basins of Spain. For example, below the exploited thenardite beds Gypsum + Halite in the Madrid (Tajo) basin, Spain, the succession in the upper part of the lower Miocene unit is characterised by Anhydrite+dolomite glauberite layers made up of a mixture of glauberite (45.8 Halite %) and halite (41.7 %), with a minor polyhalite (7.8 %), Anhydrite+dolomite Polyhalite + halite dolomite (2.1 %), and clay minerals (1.8 %) (Herrero et Anhydrite+dolomite al 2015). Polyhalite + halite The Madrid Basin is a large Tertiary intra-cratonic de- Gypsum + Halite pression that contains some of the largest fossil Polyhalite + halite sulphate and sepiolite deposits in the world. Bedded so- Anhydrite+halite dium sulphates (glauberite and thenardite) are restricted to the Lower Saline Unit, where they are associated with Polyhalite + halite anhydrite, halite, magnesite, polyhalite and minor clays. Gypsum + Halite Glauberite and thenardite are thought to have been de- posited in the most central part of a permanent saline lake. Anhydrite+halite The accumulation of thenardite might have taken place Polyhalite + halite during a stage of contraction of the lake system at the be- ginning of the middle Aragonian (middle Miocene). Anhydrite+halite 50m Polyhalite occurs as a diagenetic saline phase related both Polyhalite + halite to calcium and sodium sulphates occurrences. Both sepi- Halite+ dolomite olite and bentonite deposits are widely distributed within Dolomite peripherally in distal fan and marginal lacustrine sequenc- Figure 5. Measured section from a borehole with early diagenetic poly- es in the so-called Intermediate Unit of the Miocene halite in the Saline Unit of the Madrid Basin (After Ordonez at al., 1991) Page 5 www.saltworkconsultants.com

crystals are always associated with micritic magnesite). In its turn, the felty polyhalite may be related to skeletal A. glauberite crystals. The halite­ crystals commonly exhibit chevron-type mor­phologies. The thickness of the individ- ual layers of halite ranges from 1 to 6 cm. Similar polyhalite proportions are entrained in a number of glauberitic mineral assemblage in gypsiferous Neogene continental basins across the Iberian Peninsula, such as those of the Zaragoza (Salvany et al., 2007) or Lerín gyp- sum units (Salvany and Ortí, 1994), both occurrences are in the Ebro basin. In all cases, the polyhalite tends to be either massive or more typically a fibrous rim on large glauberite crystals. B. Polyhalite also occurs as a minor phase in some potash regions the Messinian evaporites of the Mediterranean. In the mined succession exposed in the Realmonte mine,(southern Sicily) the halite unit is approximately 400 m-thick. From the bottom to the top, it consists of irregular anhydrite and marly mudstone breccia layer up to 2 m thick followed by units A to D (Figure 6; Lugli et al., 1999). Unit A, up to 50 m thick, contains evenly lam- inated halite with anhydrite nodules and laminae passing upward to massive halite beds with irregular mudstone

NE Realmonte stratigraphy SW C. PE-26 PE-13 120m Trubi 80 Upper 40 Evaporites 0 D -40 C -80 -120 -160 -200 B -240 Halite -280 units -320 -360m Figure 7. Messinian polyhalite in the Realmonte Mine, Sicily (after A Chalk Garcia-Veigas et al., 1995. A)Photomicrograph of radial spherulitic Gypsum &marls aggregates composed of fibrous polyhalite crystals that grew within Halite a halite rock host (black) collect in Unit C (crossed polars. B) Fibrous Marls Kainite prismatic polyhalite that grew in a halite host (Unit C) and partially Basal marls replaced it. Replacement started at the halite grain boundaries (crossed A. & anhydrite polars). C) Radial spherulitic composed of fibrous polyhalite (centre of photomicrograph) that grew at the boundary among three halite grains. Unit C partially crossed polars. NE SW Realmonte structure bed some decimeters thick. Unit B (approximately 100m thick) consists of massive even layers of halite inter-bed- ? ded with thin kainite laminae, along with millimeter to centimetre-thick layers dominated by polyhalite spheru- Trubi Lower 0 200 m Gypsum lites and anhydrite laminae Figure 7; Garcia-Veigas et al., Upper Halite with 1995). It may well be that along with kainite, the layers Gypsum B. kainite beds of polyhalite spherulites are primary co-precipitates at the Figure 6. Polyhalite in the Messinian of Sicily A) Schematic showing potash bittern stage. The upper part of the succession con- potash stratigraphy in Realmonte Mine, Agrigento. B) Regional section showing structure in Realmonte region (after Decima and Wezel, 1971, tains several kainite layers up to 12 m thick. The 70–80 1973; Lugli, 1999). m thick unit C, consists of halite 10 to 20 cm thick lay-

Page 6 www.saltworkconsultants.com

ers separated by irregular mud laminae and it too contains Potash evaporites of the Carpathian Foredeep host an in- minor polyhalite and anhydrite. Unit D, up to 60 m thick, teresting sulphates group that includes about 20 sulphate begins with a grey anhydrite-rich mudstone passing to an evaporite minerals. Exploited potash deposits of the fore- anhydrite laminate sequence, followed by halite millime- deep are composed of kainite, langbeinite, kainite–lang- tre- to centimetre-thick layers intercalated with anhydrite beinite, sylvinite, polyhalite and carnallite rocks with layers laminae and decimetre-thick halite beds. of rock salt or interbedded clays and rock salt. In the areas of salt-bearing breccia, a polyhalite–anhydrite layer occurs Lugli et al. (1999) proposed that these lithologies, includ- along the contact with the potash salts bed. Halite, lang- ing the early diagenetic polyhalite, reflect the shallowing beinite and kainite dominated targeted ore levels in these and the desiccation of the evaporitic basin resulting from a potash deposits. Kieserite, polyhalite, anhydrite, sylvite and possible combination of factors: (1) uplift of the basin floor carnallite were present in smaller but significant quantities. by thrust activity, (2) simple evaporitic drawdown and (3) a These deposits, once a source of sulphate of potash, are no basin-wide drop of the Mediterranean sea level. longer mined. Polyhalite is also common as a potash contributor along A study of sulphur isotopic composition of 10 of the sul- with, in the highly deformed bittern series in the Bad- phate minerals from the Kalush-Holyh and Stebnyk pot- enian (Middle Miocene) ores of the Carpathian Foredeep Figure 0 250 km Estonia 8a). These beds are highly distort- A. Russia ed and host former potash mines Sweden Latvia Lithuania 50 km extracting a kainite-langbeinite Russia EAS Belarus ore target (Figure 8b). These pot- y

n

a T

m Poland ash-entraining salt deposits occur r E e UROPEA in western Ukraine within two G Ukraine Czech Lviv structural terranes: 1) Carpathian Republic Slovakia Moldova Lacko Foredeep (rock and potash salt) Austria Dobromil Hungary Romania N and (II) Transcarpathian trough Drohobych PL A (rock salt) (Figure 8a). Deposits Stebnyk TF

E Transcarpathian basin O differ in the thickness and lithol- N R Bolekhiv E Bilche-Volytsya Unit OUTER M ogy, depending on the regional Dolyna Kalush

C

tectonic location (Czapowski et O Sambir Unit Ivano-Frankivsk I T RANSCA Starunia

al., 2009). In the Ukrainian part M Boryslav-Pokuttya Unit CARP of Carpathian Foredeep, three Lanchyn Kolomyia T Carpathian flysch R Delyatyn main tectonic zones are distin- O R U PA A G T THIANS guished (Figure 8a): (I) outer Potash salts H HIA Chernivtsi N Kosiv zone (Bilche-Volytsya Unit), in Salt-works (brine recovery) Solotvino which the Miocene molasse de- Salt mines posits overlie the Mesozoic plat- cross section form basement discordantly at a depth of 10-200 m, and in the SW NE foredeep they subsided under the CarpathianForedeep Carpathians overthrust of the Sambir zone and Boryslav-Pokuttya Zone Sambir Zone are at depths of 1.2-2.2 km (Bu- kowski and Czapowski, 2009); 1 km Hryniv et al., 2007); (II) central 1 km zone (Sambir Unit), in which the Miocene deposits were overthrust rock salt some 8-12 km onto the external Kosiv Beds Carpathian ysch kainite-langbeinite part of the Foredeep deposits of Beds Tyras Beds (salt breccia) Regional nappe salt breccia with sylvite Badenian the external zone occur at depths Vorotyshcha Local nappe Eggenburgian Stebnyk Beds of 1.0-2.2 km; (III) internal zone B. subsalt siliciclastic Overthrust (Boryslav-Pokuttya Unit), where with olistoliths Miocene deposits were overthrust Figure 8. Potash in Carpathian Foredeep, Ukraine. A) Distribution of potash and rock salt deposits (red areas, salt mines marked by green circles, old saltworks (brine concentration since 17th Century atop the Sambir Nappe zone by grey circles) plotted on a background of the regional geological structure of western Ukraine. across a distance of some 25 km B) Geological cross-section of two time-separate zones in Carpathian Foredeep near Stebnyk (after (Hryniv et al., 2007). Bukowski and Czapowski, 2009; Hryniv et al., 2007; Koriń, 1994). Page 7 www.saltworkconsultants.com

Mineralogy Locality d34S (‰) CD 3) inflow of surface waters containing sulphates enriched Stebnyk region in light sulphur isotopes due to pyrite oxidation. Accord- Clayey polyhalite rock Stebnyk mine 16.73 ingly, the observed sulphur isotopic composition of miner- Red polyhalite rock Stebnyk mine 15.28 als from these potash deposits demonstrates the depletion Polyhalite from polyhalite-anhydrite beds Stebnyk mine 16.48 of the original marine brines and continual inflow of new Polyhalite from polyhalite-anhydrite beds Stebnyk mine 16.60 (concentrated) seawater and later meteoric access. The pre- ponderance of lighter sulphur isotopic values recorded in Gypsum replacing polyhalite from polyhalite-an- Stebnyk mine 15.99 hydrite beds the Stebnyk deposit can be explained by a more intensive Anhydrite from polyhalite-anhydrite beds Stebnyk mine 16.11 inflow of surface waters from the Carpathian nappes or Anhydrite from polyhalite-anhydrite beds Stebnyk mine 15.75 by the oxidation of a part of the pyrite hosted in the sed- iments. Whatever the case, it seems that once again poly- Kalush-Holyn region halite is an early diagenetic mineral. Red polyhalite rock Holyn mine 17.18 Polyhalite from polyhalite-anhydrite beds Dombrovo quarry 17.54 Permian polyhalite Polyhalite from polyhalite-anhydrite beds Sivka-Kalushska mine 17.36 Permian polyhalite deposits are much more impressive in Anhydrite from polyhalite-anhydrite beds Sivka-Kalushska mine 15.77 terms of volume and extent, compared to the Neogene, and Anhydrite from polyhalite-anhydrite beds Holyn mine 16.95 are exemplified by massive occurrences in the USA and Anhydrite from polyhalite-anhydrite beds Dombrovo quarry 15.94 Europe Anhydrite with carnallite from salt breccia Holyn mine 17.88 Basal anhydrite Sivka-Kalushska mine 21.00 Permian polyhalite in West Texas and New Kainite Sivka-Kalushska mine 15.93 Mexico Kainite Sivka-Kalushska mine 16.18 Polyhalite deposits are by far the most abundant, most Kainite Holyn mine 17.38 numerous, and widespread of all potash mineral occur- Langbeinite Dombrovo quarry 15.95 rences in the Delaware Basin of Texas and New Mexico Langbeinite Dombrovo quarry 16.27 ( Jones 1972; Lowenstein, 1988; Harville and Fritz, 1986). Langbeinite Dombrovo quarry 17.17 However, langbeinite and sylvite are the economically im- portant potash minerals and have been the focus of many Kieserite Dombrovo quarry 16.15 studies, rather than polyhalite documentation (Figure 9a). Leonite Dombrovo quarry 16.57 Permian polyhalite in the Delaware Basin occurs both as Picromerite Dombrovo quarry 18.22 massive and disseminated deposits in anhydrite and salt Bloedite Dombrovo quarry 16.77 beds and less often in clay beds. Typically, massive depos- Bloedite Dombrovo quarry 17.43 its and all veins and lenses are composed predominantly Syngenite Dombrovo quarry 14.73 of polyhalite, in distinctly compact units that have sharp, Gypsum Dombrovo quarry 15.41 clear-cut outlines. Disseminated deposits generally are less Gypsum Dombrovo quarry 15.96 defined, shapeless bodiesof spherules as -parallel

34 growths in a host rock, chiefly in halite. Disseminated oc- Table 3. d SCDvalues of sulphate minerals from the Kalush–Holyn and Stebnyk potash deposits of the Carpathian Foredeep (From Hryniv et currences are many times more numerous than the massive al., 2007). deposits, but the amount of polyhalite present is minor in comparison with that present in most massive deposits in ash deposits shows that only the basal Ca-sulphates (anhy- d34 anhydrite beds. drite) from the Kalush-Holyn potash deposits has SCD values typical of Neogene marine evaporites (+21.0‰; Massive polyhalite occurrences outline a crude oval- Hryniv et al., 2007). Potash minerals related to the ore-as- shaped area in the basin, extending over a region about sociations in the deposits (polyhalite, anhydrite, kainite, 325 km long and 220 km wide, covering practically the d34 langbeinite and kieserite) show SCD values from +15.28 whole southern half of the area between the Pecos River ‰ to +17.54‰, while weathering zone minerals (pi- and the eastern limit of salt in the Ochoa Series (Figure cromerite, leonite, bloedite, syngenite and gypsum) in the 9a). Occurrences range stratigraphically from low in the Dombrovo Quarry show values ranging from +14.73‰ to Tansill Formation (upper part of Guadalupe Series) in +18.22‰ (Table 3). the North-western shelf to near the middle of the Rustler Formation in the north-east corner of the Delaware basin According to Hyrniv et al. (2007) the recorded depletion (Figure 9b). Polyhalite beds reach their highest number and of sulphur isotopic composition of the salt minerals in the size in the Salado Formation (Ochoan), where they have a Ukranian potash deposits (and their weathering zone) was wide distribution over much of the Delaware and Midland probably caused by one or more factors as follow: 1) bac- basins and adjacent platform and shelf areas (Figure 9). terial reduction of sulphate, 2) effect of crystallisation and In the Salado Fm., thick clay seams occur as basal strata Page 8 www.saltworkconsultants.com

105° 104° 103° 102° 101° 100° of the Salado Formation. Sections 35° O with layered halite and polyhalite Pecos River Canyon K L cover areas of 95,000 km2 and A 70,000 km2, respectively ( Jones TEXAS NEW 1972). MEXICO 34° Portales Massive polyhalite units are typ- Extent of halite ically compact and fine-grained, Lubbock exhibiting a variety of colours (grey Roswell to red) and textures (irregular to layered to laminated and fibrous to equicrystalline prismatic). Signifi- 33° cant volumes are replacements of anhydrite beds, and although they may have almost any shape, most

Carlsbad tend to be lenticular to sheet-like masses that spread out along the 32° bedding and replace practically the Midland entire section of anhydrite. Poly- San Angelo Extent of sylvite- and halite units in the McNutt Potash langbeinite-bearing zone, east of Carlsbad, have later- units in halite al continuities sufficient to act as Van Horn 31° marker beds, which separate and Extent of polyhalite define layering in the sylvite-lang- in halite beinite ore zones (Figure 10). MEXICOU.S.A. Alpine Pecos River As a general rule, sheet-like to 100 km crudely tabular polyhalite bodies A. occur in anhydrite layers where stacked polyhalite units are a few centimetres to a metre thick. De- NW Northwestern Shelf Reef Zone SE Eddy Co. Delaware Basin Lea Co.

Chaves Co. Eddy Co. posits that are more irregular in Rustler Formation shape occur mostly in thicker beds Limits of soluble of anhydrite (>1m). IN most cases

Ochoan potassium minerals tash depo the polyhalite is pseudomorphous Bed ded po sits Limits of polyhalite after growth-aligned subaqueous T occurrence ansil Ya l and nodular gypsum or nodular tes anhydrite beds (Figure 11).

Salado Formation Practically all the deposits en- Capitan

Permian-Guadalupian Reef close residual strips and irregular Metres 300 Castile Anhydrite remnants of magnesitic anhydrite, 200 which are mottled and streaked 100 with halitic and anhydritic pseu- ell Ca B nyo 0 n F domorphs after gypsum. Com- 10 20 ormation B. km monly polyhalite crystals and mul- tigrain aggregates project into the Figure 9. Distribution of massive potash deposits in evaporites of the Ochoan (Salado) Series in Texas magnesitic anhydrite remnants and New Mexico A) polyhalite has a wide extent, while the economic sylvite-langbeinite layers are restricted to a smaller area known as the McNutt Zone located to the east of Carlsbad (Jones 1972). either as elongate crystals and B) stratigraphic section in southeastern New Mexico showing distribution of bedded sylvite-langbeinite veinlike tongues or as aggregates mineralization with the polyhalite interval of the Salado Fm (after Bates, 1969). having scalloped margins convex toward anhydrite. that underlie massive polyhalite/anhydrite beds (Harville In many places in the Carlsbad district and nearby parts and Fritz, 1986; Lowenstein, 1988). By virtue of the wide of the north-western shelf, many of the massive polyha- extent and number of massive deposits, polyhalite ranks lite deposits grade laterally to an anhydritic hartsalz unit next to halite and anhydrite among the major constituents with ore grade levels of sylvite. This is the area known as Page 9 www.saltworkconsultants.com

Metres below Potash zone -mostly with sylvite-langbeinite Rustler-Salado Unit and halite-clay gangue contact Polyhalite 150 Anhydrite Voca Triste Sand OCHOAN 160 Halite and shaly halite

Dewey Lake Shaly, sandy halite 170 11th Ore Zone (mostly carnallite, minor sylvite and leonite - noncommercial) Rustler Marker bed No 121

103 180 Marker bed No 121 108 Upper

Member Marker bed No 121 Voca T. 190 10th Ore Zone (sylvinite, high clay content of 6-7%, was 2nd best ore target in district) Union Marker bed No 120

McNutt 9th Ore Zone (carnallite, kieserite, minor sylvite - non commercial) Member 126 Marker bed No 121 Salado 200 Marker bed No 122 128 8th Ore Zone (moderate reserves, high clay, mined) 130 210 Union Anhydrite 7th Ore Zone (sylvite, moderate reserves, moderate clay) Cowden

Lower Member Lower 6th Ore Zone (carnallite, kieserite, minor sylvite - noncommercial) 220 5th Ore Zone (sylvite and langbeinite, trace clay - 1%)

Marker bed No 123 230 Castile Marker bed No 124 4th Ore Zone (langbeinite and sylvite, mixed ore, main source of langbeinite) 240 3rd Ore Zone (sylvite, historically was considered 3rd-ranked ore taget) 2nd Ore Zone (carnallite, kieserite, minor sylvite, noncommercial) 250 Marker bed No 125 Halititic Dolomitic 1st Ore Zone (sylvite, historically was principal ore target, now depleted) Marker bed No 126 Anhydritic Sandstone/Shale 260

Figure 10. Stratigraphy of the McNutt Potash zone in the Salado Fm, showing polyhalite marker beds, focused on the main ore zone near Hobbs New Mexico (comments on from Griswold, 1982). the McNutt Member or the McNutt potash zone (Figures All units are interpreted as mostly marine-brine dominat- 9a, 10). The change from polyhalite to hartsalz coincides ed units precipitated by evaporation of massive volumes with a shift from unmineralized to sylvinitic rock peppered of brines fed by marine seepage or periodic overflows of with sparse grains and veinlets of carnallite and other mag- the Permian ocean water. The upper cap to Type II cycles nesium-bearing bittern minerals, such as langbeinite and influenced by inflows of continental groundwater (Figure polyhalite. 11). In 1988, Lowenstein recognised two types of metre-scale A basal mudstone grades upsection into anhydrite-poly- depositional cycles (Type I and Type II) within the halite that is commonly laminated. Laminae are defined McNutt Potash Zone (Figure 11). Both cycles record pro- by couplets of anhydrite or polyhalite separated by magne- gressive drawdown and concentration of brine in a shallow, site-rich mud (Figure 12a-c). The most significant feature marginal marine drawdown basin. "Type I" cycles have a of the anhydrite/polyhalite interval is the large number of base of carbonate-siliciclastic mudstone, overlain by an- crystal outlines that occur in the anhydrite-polyhalite lami- hydrite-polyhalite that is pseudomorphous after primary nae. These crystals are now composed of anhydrite, polyha- bedded gypsum. This, in turn, is overlain by bedded halite lite, halite, or sylvite but are all interpreted as replacement and capped by muddy halite. Lowenstein (1988) conclud- pseudomorphs after primary gypsum because of their close ed the McNutt Zone of the Salado Formation consists en- similarity to typical bottom-nucleated subaqueous gypsum tirely of these two types of metre-scale sequences, variably ,s such as "swallow-tail twins" (Figure 11). In stacked one upon another (Figure 11). some occurrences, the polyhalite is forming early diagenet- ic spherules in magnesite layers (Figure 13a). Elsewhere Page 10 www.saltworkconsultants.com

places, rippled gypsum beds are replaced by polyhalite and anhydrite. Syndepositional brine reflux likely drove re- placement of subaqueous gypsum by anhydrite-polyhalite, in a fashion similar to that described by Hovorka (1992) for halite replacing growth-aligned gypsum. At the microscopic scale, it is evident that polyhalite forms as a replacement (Figure 13). One of the most common modes of occurrence across the Salado Formation is as co- alescing spherules growing in relatively undisturbed mag- nesite layers (Figure 13a). Elsewhere, coarser mm-scale polyhalite prisms have poikilotopically enclosed anhydrite crystals (Figure 13b). Felted fibrous polyhalite also sur- rounds euhedral halite (Figure 13c) or forms a replacement rim to halite in the langbeinite-sylvite ore layers (Figure 13d). "Type II" cycles, lacking the basal mudstone and polyha- lite/anhydrite beds, occur between Type I cycles and con- tain additional halite units (with thinly layered polyhalite) overlain transitionally by muddy halite (also with dispersed polyhalite). Complete -upward Type I and Type II cycles record a temporal evolution of depositional environ- ment from a shallow saline lake to an ephemeral salt-pan- saline mudflat complex. The uppermost muddy halite unit interpreted as a continental-dominated sequence, sourced Figure 11. Vertical cyclic sequences in the McNutt Potash Zone of the by meteoric inflow from surrounding land areas that mixed Salado Formation, with diagnostic sedimentary structures and textures with variable amounts of seawater, either from residual pore with polyhalite present at all levels,but especially common as replacement waters or introduced into the Salado Basin by seepage. in the anhydrite beds toward the base of the cycles. Includes interpreted inflow waters that were present during formation of each layer (after Periodic invasions of seawater best explain the vertical Lowenstein, 1988). stacking of Type I cycles in the Salado basin, perhaps coin- cident with eustatic sea-level rises (Lowenstein, 1988). The polyhalite directly replaces bottom-nucleated subaqueous continental-dominated upper parts of Type I and II cycles gypsum or halite (Figures 12a, c, ), while yet elsewhere it formed during intervening periods of eustatic sea-level fall grows as spherular clusters in halite that already has pseu- and low stand when nonmarine waters exerted more in- domorphed aligned gypsum crystals Figure 12c). In other Polyhalite replacing growth-aligned gypsum

Spherular polyhalite in halite psuedomorphs Polyhalite replacing replacing growth- halite pseudomorphs aligned gypsum of growth-aligned gypsum with disrupted magnesite drapes Anhydrite layer

Magnesite between polyhalite-anhydrite layers A. 398m B. 177m C. 397m Borehole 4 Borehole 6 Borehole 4 Figure 12. Examples of polyhalite replacing gypsum in slabbed core faces (core images extracted and recompiled from Schaller and Henderson, 1932. Note the white magnesite layers and the intimate tie to anhydrite layering.

Page 11 www.saltworkconsultants.com

cent part of Russia (Figure 14). In addition, K-Mg chlorides are M found locally both in Poland and Russia. The K-Mg salts originated during the last stages of chloride accumulation within small, active- ly subsiding isolated salt basins of the salina type, which were proba- A. B. bly tectonically controlled. S The paragenetic sequence in one S polyhalite (Zdrada) deposit in the Zechstein of Poland was the result of a very early - penecontempora- neous polyhalitisation of anhydrite that had already pseudomorphed gypsum, much as is seen in the S Delaware basin (Peryt et al. 1998). C. D. There polyhalite formed by alter- ing anhydrite during crossflows of Figure 13. Petrography of polyhalite (after Schaller and Henderson, 1932). A) Spherulitic fibrous polyhalite growing in a magnesite background (black) this is the typical occurrence of polyhalite in concentrated brines that were also magnesite layers. B) Prismatic polyhalite poikilotopically enclosing finer-crystals of anhydrite in a responsible for potash deposition background of halite. C) Halite crystals unclosed in fine-grained felted fibrous polyhalite. D) Polyhalite in local salt basins, while the sul- fibres rimming halite crystals with growth entry along halite cleavage planes, in a background of sylvite. phate-rich brines supplied by the dissolution of emergent parts of fluence on the brine chemistry. According to Lowenstein the sulfate platform (Peryt et al. 1998). (1988), the maximum time interval between major marine incursions averages 100,000 years. The layered nature of The timing of the polyhalitisation can be inferred from a the polyhalite re- HST placement implies PZ4 TST A1g K Na1 Na1 Baltic Sea that this occurred PZ3 Hiatus LST A1d in each eustatic cy- Mieroszyno Chlapowo PZ2 HST deposit cle, that is, the re- Ca1 deposit Permian Zechstein PZ1 TST T1 placement was an LST TST Zc1 integral part of the Lithuania and Swarzewo Poland Kalingrad region anhydrite breccia deposit eogenetic hydrol- A1g Upper anhydrite potash Na1 Oldest halite rock salt Priegliaus ogy and was not a A1d Lower anhydrite burial diagenetic anhydrite Ca1 Zechstein limestone Naujosios Akmenes Zdrada Puck Bay carbonate T1 Kupferschiefer Sasnavos deposit (mesogenetic) pro- marl, claystone Zc1 Basal conglomerate Kalvarijos 2 km A. OLDEST HALTE cess. siliciclastics Borehole Sample Depth (m) 37Cl (‰) Zdrada IG3 3A 804.3 0.2 Permian 7 784.0 0 Baltic Sea Zdrada IG6 3 941.5 -0.1 Polyhalite in 12 904.3 0.2 Zdrada IG8 1 981.1 0 Poland and 4 947.7 0.29 8 916.7 0.4 Russia 16 846.6 0.2 According to Peryt 12A 806.2 0 et al., 2005 (and 20 784.7 0.1 B. references there- Platform Slope Basin Polyhalite deposit Puck Bay in) there are four Polyhalite in Oldest Halite K-chloride salt in the polyhalite deposits Polyhalite in Lower Anhydrite Oldest Halite LOWER ANHYDRITE 2 km in the Zechstein of Polyhalite in Oldest Halite No polyhalite C. and Lower Anhydrite northern Poland, and more than ten Figure 14. Polyhalite deposits in the Zechstein of Poland (after Peryt et al., 2005). A) Stratigraphy of the Zechstein in the polyhalite-bearing Peribaltic Basin. B) isotope data for the Oldest Halite in the Zdrada area. C) Paleogeography and occurrence of potassic salts in the Lower Anhydrite and Oldest Halite in the region west of Puck Bay. areas in the adja- Page 12 www.saltworkconsultants.com

12.0 from the Boulby Mine was around 1 Mt/yr of refined KCl product Zechstein, Poland and 0.6 Mt of road salt (Kemp et al., 2016). Polyhalite beds in the proposed York (Whitehall) mine are considered to be so high grade Anhydrite that they can be mined and mar-

CDT keted as SOPM fertiliser with no S 11.0

34 processing except crushing and

δ sizing (Kemp et al., 2016). Anhydrite Five evaporite cycles (EZ1-EZ5) plus polyhalite are developed in the northwestern corner of the main Permian Zech- stein basin where it comes onshore 10.0 in the UK between Teesside and -20.5 -20 -19.5 -19 -18.5 Lincolnshire (Table 2, Figures 16, 17). δ18O SMOW The relationship between the evap- Figure15. Oxygen and sulphur isotope compositions of sulphate evaporites of the Zdrada platform orite sequence in the main Zech- (Peryt et el., 1998). stein basin and its onshore, later- S-O isotope crossplot (Figure 15; Peryt et al., 1998). The al gradation into shelf and then isotopic compositions of sulphate evaporites indicate that semi-continental clastic strata was described by Smith aet marine solutions were the only source of sulphate ions al., (1986). Potash salts are known from cycles EZ2, EZ3, supplied to the Zechstein basin. The more negative oxygen and EZ4, and Britain’s only potash producer, the Boulby values associated with the polyhalite compared to its anhy- mine, exploits sylvite from the EZ3 Boulby Potash Mem- drite precursor indicates somewhat warmer solutions that ber. Sylvite-bearing horizons are also known in the EZ2 drove the conversion to polyhalite. These solutions were cycle, but the key potash resource therein is polyhalite, first more saline than those driving the initial shallow anhydri- discovered in 1939 in the E2 oil exploration hole at Esk- tisation that replaced platform gypsum by a reaction with dale, (Stewart, 1949). The only known occurrence refluxing brines. of potentially economic volumes of polyhalite in the UK is in the EZ2 Fordon (Evaporite) Formation in this area.

Polyhalite in the Zech- 1°0'0''W 0°30'0''W 0°0'0' ' stein of the UK Edinburgh Polyhalite in the Boulby Mine Teesport and the proposed York mine Redcar both occur within the Permian Boulby Potash Mine London Fordon Formation, of the 2nd Saltburn-by-the-Sea Zechstein cycle (Z2) in north- east (Figure 16; Table 4; Whitby York Potash borehole Historical borehole Stewart 1963; Smith et al., 1986; 54°30'0''N York Potash Mine York Potash Kemp et al. 2016). Although (proposed) SM2 SM7 SM1 5km initially discovered in 1939, the Eskdsale 3 Robi n SM11 SM3 Hood’s Bay deeper, polyhalite-bearing For- Eskdsale 1 SM9 Stoupe Beck 1 don (Evaporite) Formation was North Yo rk Moor s SM6 National Park largely overlooked until recent- SM4 ly. ICL-UK operations at the Boulby Mine have largely de- Scarborough pleted the sylvinite target in the Boulby Potash Member, so the Kirkbymoorside mine is now transitioning into 54°15'0''N Pickering polyhalite extraction from the Fordon (Evaporite) Formation Figure 16. Map indicating the York Potash area of interest, the location of the Boulby and York (Table 4). The historical output (Whitehall) mines, and sites of various exploration and historical boreholes (after Kemp et al., 2016). Page 13 www.saltworkconsultants.com

Zechstein North East England onshore lithostratigraphy thickness. The Upper subcycle formed in uniformly shal- cycle low-water conditions with no clear distinction between EZ5 Eskdale Group Littlebeck (Anhydrite) Fm. shelf and basin. It hosts a persistent sylvite-bearing hori- Sleights (Siltstone) Fm. zon known as the Gough Seam. Colter and Reed (1980) EZ4 Staintondale Sneaton (Halite) Fm. Sneaton Potash Mbr Group Sherburn (Anhydrite) Fm. showed that Stewart’s mineral zones could be projected Upgang (Limestone) Fm. far beyond the Fordon borehole and were recognisable Carnallitic Marl Fm. throughout much of the British section of the North Sea EZ3 Teesside Boulby (Halite) Fm. Boulby Potash Mbr basin (Doornenbal and Stevenson, 2010). Group Billingham (Anhydrite) Fm. Brotherton (Mg limestone) Fm. The description of mineral zones at Eskdale and Fordon Grauer Salzton by Stewart (1949, 1963) relate to boreholes drilled through EZ2 Aislaby Fordon (Evaporite) Fm. Polyhalite unit the shelf and basin, respectively. The precise correlation of Group Kirkham Abbey (Limestone) Fm. the polyhalite- bearing sulfate deposits between these two EZ1 Don Group Hayton (Anhydrite) Fm. environments, or zones, remains ambiguous (Kemp et al., Cadeby (Limestone) Fm. Marl Slate Fm. 2016). At present, the polyhalite deposit is referred to as the Shelf seam in the Shelf zone, and the Basin seam in Table 4. Permian lithostratigraphical units in northeastern England (after Smith et al., 1986). EZ = English Zechstein units. the Basin zone, with a Transition zone across the ramp and in its vicinity where great thicknesses of polyhalite and Mineral zonation in the Fordon (Evaporite) Formation anhydrite occur with varying amounts of early diagenetic, was first described in detail by Stewart (1949, 1963) from displacive halite. In borehole SM2 there was solid evidence the Eskdale and Fordon boreholes. Polyhalite was de- for overlapping Shelf and Basin seams, separated by 82 m scribed as partly primary, but mostly a replacement of syn- of “sulphatic halite”. Both the shelf and basin polyhalite depositional anhydrite. Three subcycles were recognised at seams are considered to be of mineable thickness and grade Fordon. The Lower subcycle was deposited in a basin that in their relevant sectors, averaging over 12 m in thickness still displayed considerable topographic variation from a for high-grade sections of >85% polyhalite. shallow-water shelf to a deepwater basin (Figure 17). It Kemp et al. (2016) argue that the polyhalite is almost contains no known potash occurrences. The Middle subcy- entirely secondary, resulting from replacement reactions cle, in which the polyhalite occurs, includes a large volume between freshly deposited anhydrite muds on the seabed, of basin-fill evaporites, chiefly halite, that filled accommo- with dense, bottom flowing, K-Mg-rich brines. A syl- dation space and smoothed out the shelf-basin geometry. vite-bearing bittern salt horizon is locally present near the Consequently, it shows considerable lateral variation in top of the Middle subcycle in both the Basin and the Shelf

West East Eskdale 11 SM3, SM4 Stoupe Eskdale 13 SM6 SM9 SM7, SM11 SM2 SM1 Beck1

Grauer Saltzon Gough Seam Upper subcycle Upper anhydrite Ka Ka Pasture Shelf polyhalite seam Beck Seam Basal Anhydrite Fordon Middle subcycle (Evaporite) Basin ll halite Formation 200m Black Halite Basin Polyhalite Seam Kirkham Abbey Formation Lower subcycle Not to scale Shelf zone Transition zone Basin zone ~14 km

Polyhalite Halite Anhydrite Limestone Sylvinite seam Sulphatic halite (anhydritic) Sulphatic halite (polyhalite) Ka Kalistronite

Figure 17. Conceptual geologic model for the Fordon (Evaporite) Formation in the York Potash Ltd area of interest (after Kemp et al., 2016).

Page 14 www.saltworkconsultants.com

B.

A. C. Figure 18. Polyhalite textures. A) Example of slabbed core from the potential ore zone, showing variably inclined bedding in banded polyhalite with magnesite and anhydrite. Depth 1,535 m, borehole SM3A. Scale at the side of the core tray is in 5-cm intervals (after Kemp et al., 2016). B) Boulby mine wall view of a polyhalite dome (tepee) structure illustrating the polyhalite and halite bands, a halite nucleus and apparent fill sparry halite conduits in between inclined polyhalite bands within the tepee structure and breccia, C) Schematic diagram with some thickness measurements indicated with arrows (from Abbott, 2017).

(though less commonly) and is referred to here as the Pas- crests that define most tepees (Kendall and Warren, 1987). ture Beck seam, after the borehole (also known as SM1) Instead, the domal peak tends to be a fractured and fold- where it was first cored and characterised (Figure 16). ed local anticlinal culmination. Whether one calls these anticlinal deformations domes in the polyhalite a true te- Another sylvite-bearing bittern salt horizon is more com- pee, depends on which definition of a tepee one chooses to monly present near the top of the Upper subcycle in both use. The domal features are thought to be a soft sediment Basin and Shelf and is referred to here as the Gough seam; deformation features, formed via polyhalite dewatering, described in the SM4 borehole, where it was first cored coupled with penecontemporaneous precipitation of halite and characterised as containing relatively high-grade syl- in opening fractures and below anticline crests in shallow vite. It is not clear why this and the Pasture Beck potash burial. Deformation was driven by fluid crossflows and -es seam are so localised and patchy in distribution, but they capes, as anhydrite converted to polyhalite. may result from bittern brine pools of limited area, cut off from each other as the aggrading basin filled up at the end Forming polyhalite? of each subcycle. Nowhere is the present or the past is there evidence of At an even more local scale in the polyhalite ore intervals direct primary precipitation of polyhalite. By primary, I in the Boulby Mine, there are metre-scale domal= struc- mean that to be considered a primary polyhalite, the crys- tures interpreted as a form of tepee structure (Figures 18; tals should drop out of a concentrating at-surface brine Abbott, 2017). The height of the domal-shaped structures either as bottom-nucleated or foundered brine-surface exposed in the mine workings varies between ~0.4 m and crystals. Such primary textures are widespread in gypsum 1.5 m (average = 0.9±0.1 m) and widths ranging from ~2.3 and halite units but not in polyhalite. Instead, polyhalite m to 10.5 m (average = 5.3±0.5 m). textures and isotopic signals indicate polyhalite forms via Unlike the highly deformed halokinetic flow textures in replacement of gypsum or anhydritised gypsum. the overlying sylvite of the Boulby Potash Member, it In the modern salt flats of Ojo de Liebre, we see poly- seems much of the polyhalite ore preserves mostly syn- halite replacing gypsum. Likewise, in various Tertiary la- depositional diagenetic alteration structures. Most of the custrine basins in Spain, polyhalite is found in association domal features do not show the overthrust brittle ridge with gypsum and thenardite/, and it is replacing a CaSO4 Page 15 www.saltworkconsultants.com

phase. In the Badenian marine a. Highstand carbonate - open marine connection evaporites of the Carpathian fore- s.l. deep, the polyhalite is part of the kainite-langbeinite ore sequence. It is in the Permian of the UK and West Texas and New Mexi- co that the volumes of polyhalite b. Isolated basin- basinwide & gypsum - platform • Lowstand with respect to world sealevel become sufficient for it to become saltern/mud at (unconnected) but highstand with respect to drawdown levels created by isolation of basin (gypsum b.l. a potential ore target in its own slope keep-up phase). Similar to Tucker (1991) model 1. Transition from a. to b. can deposit organic-rich right. Once again all the textural basin mesohaline source rocks. and isotopic evidence indicates polyhalitisation of anhydrite rath- c. Isolated basin - basinwide gypsum to halite saltern/mud at with local deeps (>2m-20m) • Lowstand with respect to world sealevel er than primary precipitation. But (unconnected) and lowstand with respect to possible b.l. drawdown levels in isolated basin (halite start-up this replacement is more likely to phase). Similar to Tucker (1991) model 2. be eogenetic (driven by nearsur- face hydrologies that were active in the depositional setting) rather d. Isolated basin - aggrading basinwide halite saltern/mud at with local deeps (>2m-20m) b.l. • Lowstand with respect to world sealevel (unconnected) than mesogenetic (burial). but aggrading brine level means it is a transgressive systems tract with respect to possible drawdown levels The most likely driven mechanism in the isolated basin (halite keep-up phase). was brine reflux moving highly saline seawater through shallowly e. Isolated basin - “ll and spill” transition into a humid climate creates less saline buried units of platform or ba- beds (gypsum and lacustrine carbonates) or into arid conditions marked by more saline salt beds (potash and other bittern salts) sinal gypsum and anhydrite. This b.l. • Still a low amplitude lowstand with respect to sea shallow subsurface emplacement level (sea has not yet breached isolating barrier) but a highstand with respect to possible drawdown levels in occurred while the gypsum anhy- the basin. drite was still permeable, and so allowed the preservation of pris- f. Marine connected -Highstand aggraded platform tine texture (pseudomorphs) of s.l. • Highstand with respect to sea level and prole has the CaSO4 precursors. attained this geometry via marine carbonate ll. Tectonic or halokinetic withdrawal may have generated and maintained a new rimmed-shelf/shoal prole during this Polyhalitisation of basinal and time. platform gypsum units in the me- ga-sulphate stages of a saline giant Figure 19. Times and positions of hydrologies suitable for polyhalitisation via brine reflux is indicated are driven by time separate hydrol- by the position of blue arrows in holomictic gypsum saturation hydrologies, in a hydrograpically ogies, tied to the changing brine isolated marine-seepage saline giant. The basinal sulphates are converted to polyhalite at time b, while the platform sulphates are converted at time c. The sequence stratigraphic model driving the levels in the drawndown basin seepage positioning is based on the sequence stratigraphic model of Warren 2016. (Figure 19; Warren, 2016). Ma- rine-derived brine reflux through platform and basin polyhalite units contains evidence of basinal occurs during maximum drawdown in both lamination and subaqueous shallow water deposition the mega-sulphate basin (blue arrow positions stage b in (Figure 20). In any megasulphate saline giant, the basin Figure 19), while reflux through the upper (marginal salt- brine level can oscillate between shallow and deep and, de- ern) parts of a sulphate platform is a response to a relative pending on the nature of the overlying brine column, and highstand (blue arrow positions in stage c Figure 19; War- we can deposit primary textures that are mm-cm laminates ren, 2016 - Chapter 5). The likely loss of permeability as or upward-aligned gypsum growths, or displacive nodules. one goes deeper in a sulphate platform and the associated Bottom nucleated, upward aligned gypsum crystals indi- lessening in the volume brine crossflow probably explains cate relatively stable and saturated bottom chemistries be- why there is an interval of sub-economic polyhalitic sul- neath a holomictic brine column (Chapters 1 and 2, War- phate separating the basinal from the shelfal ore zones. ren 2016). Without holomixis, brine reflux cannot occur. The sequence stratigraphic fill model also explains why the Layered and laminated gypsum sediments interlayered by patchy potash intervals are located higher in the stratigra- carbonates indicate subaqueous deposition with fluctu- phy at the "fill and spill stage" of a hyperarid climate (stage ating chemistries in the overlying column Laminites can e in Figure 19; Warren 2016). form via changes in water chemistry in a meromictic deep A drawdown model encompassing two stages of polyha- water mass (as in the modern prior to 1979), or litisation explains why much of the textures seen in the it can indicate a shallow overlying water mass subject to periodic freshening as in the salinas of southern Australia. Page 16 www.saltworkconsultants.com

Time 1. Polyhalitisation via brine re ux of basinal gypsum in the maximum drawdown in the early mega-sulphate stage the Neogen, which is also a time of evaporite inll. For re ux to occur the drawndown saltern brine body must be holomictic (see Stage b in Figure 18) of MgSO4-enriched waters. Poly- Drawndown marine-fed brine surface halite is never present in the gyp- Basal Anhydrite Fordon Re uxing brine (Evaporite) siferous units of the Messinian or Formation 200m Black Halite Badenian saline giants in the same Basin Polyhalite Seam Middle subcycle volumes we see in the Permian. Kirkham Abbey Formation Lower subcycle Then again, the extreme hyper- Not to scale arid hydrologies we see in arid Exposure and subaerial degradation Basinal shallow water gypsum climate belts across the Pangean

Time 2. Polyhalitisation via brine re ux in a platform gypsum (transgressive system tract) in the later mega-sulphate stage supercontinent are also unusual. of evaporite inll. For re ux to occur the drawndown saltern brine body must be holomictic (see Stage c in Figure 18) But the seawater chemistry was not too different to that of today Marine -fed brine surface (Lowenstein et al., 2005). Ka Shelf polyhalite seam Holomictic Fordon Basal Anhydrite Zone of lesser penetration brine Middle subcycle (Evaporite) References of the re uxing brines Formation 200m Black Halite Abbott, S., 2016, Depositional Basin Polyhalite Seam Kirkham Abbey Formation architecture and facies variability Lower subcycle Not to scale in anhydrite and polyhalite se-

Platform gypsum Deepwater gypsum quences: a multi-scale study of the

~14 km Jurassic (Weald Basin, Brightling Mine) and Permian (Zechstein Polyhalite Halite Anhydrite Limestone Ka Kalistronite Basin, Boulby Mine) of the UK: Sulphatic halite (anhydritic) Sulphatic halite (polyhalite) Doctoral thesis, Imperial College FigureFigure 20. 20. The The drawdown-controlled drawdown-controlled position of re uxposition brines of in reflux the two mainbrines stages in ofthe deposition two main in a megasulphatestages of deposition basin. in London. a megasulphate basin. Andriyasova, G. M., 1972, Poly- If the layer and laminate gypsum.anhdrite is interlayered halite in Kara Bogaz: Khim. Geol. with units of bottom-aligned gypsum or its anhydritised Nauk, v. 3, p. 45-49. "ghosts," as in west Texas and Poland, then the depositing Barbarick, K. A., 1994, Polyhalite application to sor- waters in both units were shallow. ghum-sudangrass and leaching in soil columns: Soil Sci- We can now take this reflux model for polyhalitisation and ence, v. 151, p. 159-166. explain why the two polyhalite ore seams in the Forden Bates, R. L., 1969, Potash Minerals: Geology of the indus- Evaporite Formation are separated by a low quality "sul- trial rocks and minerals: New York, Dover Publ., 370-385 phatic halite-anhydrite" unit (Figure 20). At time 1 the and 439-440 p. basin is at its maximum lowstand and dense reflux brines are sinking into the basinal gypsum units. Water depths Braitsch, O., 1971, Salt Deposits: Their Origin and Com- below the holomictic brine mass in the basin lows are rela- positions: New York, Springer-Verlag, 297 p. tively shallow. At time two the brine levels in the basin are Bukowski, K., G. Czapowski, S. Karoli, and M. Babel, 2007, much higher, and a gypsum platform is prograding into the Sedimentology and geochemistry of the Middle Miocene basin. Water depths above the platform are shallow, while (Badenian) salt-bearing succession from East Slovakian they are deep in the basin centre. When the water column Basin (Zbudza Formation): Geological Society, London, is holomictic, brine reflux is occurring across the platform Special Publications, v. 285, p. 247-264. and out into the basin. However descending brines cannot Colter, V. S., and G. E. Reed, 1980, Zechstein 2 Fordon penetrate into all parts of the platform due to compaction Evaporites of the Atwick No. 1 borehole, surrounding ar- and earlier reflux of halite- saturated cements. Brines must eas of N.E. England and the adjacent southern North Sea, pass beyond halite saturation to reach polyhalite (Figure in H. Fuchtbauer, and T. M. Peryt, eds., The Zechstein Ba- 1). This early loss of permeability created a core of less al- sin with Emphasis on Carbonates: Stuttgart, E Schweizer- tered anhydrite below the polyhalite replacement interval. barts scheverlagsbuchhandlung, p. 115-129. But we must now ask, why did polyhalitisation of large Czapowski, G., K. Bukowski, and K. Poborska-Młynarska, parts of sulphate platforms reach its zenith in the Permian. 2009, Miocene rock and potash salts of West Ukraine: A pseudomorphing process with a halite-gypsum focus is Field geological-mining seminar of the Polish Salt Min- seen throughout the rock record (Chapter 5, 7; Warren, ing Society. Geologia (Przegląd Solny 2009), Wyd. AGH, 2016). But the volumes of polyhalite we see in the Permian Kraków, 35, 3: 479-490. (In Polish, English summary). saline giants are different to the much smaller volumes of Page 17 www.saltworkconsultants.com

Decima, A., and F. Wezel, 1973, Late Miocene evaporites Geology of Saline Deposits, v. 7: Paris, UNESCO Earth of the central Sicilian Basin; Italy: Initial reports of the Science Series, p. 191-201. Deep Sea Drilling Project, v. 13, p. 1234-1240. Kemp, S. J., F. W. Smith, D. Wagner, I. Mounteney, C. P. Decima, A., and F. C. Wezel, 1971, Osservazioni sulle Bell, C. J. Milne, C. J. B. Gowing, and C. J. B. Pottas, 2016, evaporiti messiniane della Sicilia centromeridionale: Riv- An Improved Approach to Characterize Potash-Bearing ista Mineraria Siciliana, v. 130–132, p. 172–187. Evaporite Deposits, Evidenced in , United Demicco, R. V., T. K. Lowenstein, L. A. Hardie, and R. Kingdom: Economic Geology, v. 111, p. 719-742. J. Spencer, 2005, Model of seawater composition for the Kendall, C. G. S. C., and J. K. Warren, 1987, A review of Phanerozoic: Geology, v. 33, p. 877-880. the origin and setting of tepees and their associated fabrics: Dong, Z., P. Lv, G. Qian, X. Xia, Y. Zhao, and G. Mu, Sedimentology, v. 34, p. 1007-1027. 2012, Research progress in China's Lop Nur: Earth-Sci- Koriń, S. S., 1994, Geological outline of Miocene salt-bear- ence Reviews, v. 111, p. 142-153. ing formations of the Ukrainian fore-Carpathian area (In Doornenbal, H., and A. Stevenson, 2010, Petroleum geo- Polish, English summary): Przegląd Geologiczny, v. 42, p. logical atlas of the southern Permian Basin Area: Houten, 744-747. Netherlands, EAGE Publications, 342 p. Lowenstein, T. K., 1988, Origin of depositional cycles in a Garcia-Veigas, J., F. Orti, L. Rosell, C. Ayora, R. J. M., and Permian ''saline giant''; the Salado (McNutt Zone) evap- S. Lugli, 1995, The Messinian salt of the Mediterranean: orites of New Mexico and Texas: Geological Society of geochemical study of the salt from the central Sicily Basin America Bulletin, v. 100, p. 592-608. and comparison with the Lorca Basin (Spain): Bulletin de Lowenstein, T. K., L. A. Hardie, M. N. Timofeeff, and R. la Societe Geologique de France, v. 166, p. 699-710. V. Demicco, 2003, Secular variation in seawater chemistry Griswold, G. B., 1982, Geologic overview of the Carlsbad and the origin of basinal brines: Geology, potash-mining district: Circular New Mexico Bureau of v. 31, p. 857-860. Mines and Mineral Resources, v. 182, p. 17-22. Lugli, S., 1999, Geology of the Realmonte salt deposit, a Harvie, C. E., J. H. Weare, L. A. Hardie, and H. P. Eug- desiccated Messinian Basin (Agrigento, Sicily): Memorie ster, 1980, Evaporation of sea water; calculated mineral se- della Societá Geologica Italiana, v. 54, p. 75-81. quences: Science, v. 208, p. 498-500. Ma, L., T. K. Lowenstein, B. Li, P. Jiang, C. Liu, J. Zhong, J. Harville, D. G., and S. J. Fritz, 1986, Modes of diagenesis Sheng, H. Qiu, and H. Wu, 2010, Hydrochemical charac- responsible for observed successions of potash evaporites teristics and brine evolution paths of Lop Nor Basin, Xin- in the Salado Formation, Delaware Basin, New Mexico: jiang Province, Western China: Applied Geochemistry, v. Journal Sedimentary Petrology, v. 56, p. 648-656. 25, p. 1770-1782. Herrero, M. J., J. I. Escavy, and B. C. Schreiber, 2015, The- Ma, L., Q. Tang, B. Li, Y. Hu, and W. Shang, 2016, Sedi- nardite after mirabilite deposits as a cool climate indicator ment characteristics and mineralogy of salt mounds linked in the geological record: lower Miocene of central Spain: to underground spring activity in the Lop Nor playa, Clim. Past, v. 11, p. 1-13. Western China: Chemie der Erde - Geochemistry, v. 77, p. 383-390. Holser, W. T., 1966, Diagenetic polyhalite in recent salt from Baja California: American Mineralogist, v. 51, p. 99- Mello, S. d. C., F. J. Pierce, R. Tonhati, G. S. Almeida, D. 109. D. Neto, and K. Pavuluri, 2018, Potato Response to Poly- halite as a Potassium Source Fertilizer in Brazil: Yield and Hovorka, S. D., 1992, Halite pseudomorphs after gypsum Quality: HortScience, v. 53, p. 373-379. in bedded anhydrite; clue to gypsum-anhydrite relation- ships: Journal of Sedimentary Petrology, v. 62, p. 1098- Ordóñez, S., J. P. Calvo, M. A. García del Cura, A. M. 1111. Alonso-Zarza, and M. Hoyos, 1991, Sedimentology of Sodium Sulphate Deposits and Special Clays from the Hryniv, S., J. Parafiniuk, and T. M. Peryt, 2007, Sulphur Tertiary Madrid Basin (Spain): Lacustrine Facies Analysis, isotopic composition of K Mg sulphates of the Miocene Blackwell Publishing Ltd., 39-55 p. evaporites of the Carpathian Foredeep, Ukraine: Geolog- ical Society, London, Special Publications, v. 285, p. 265- Pavuluri, K., Z. Malley, M. K. Mzimbiri, T. D. Lewis, and 273. R. Meakin, 2017, Evaluation of polyhalite in comparison to muriate of potash for corn grain yield in the Southern Jones, C. L., 1972, Permian Basin potash deposits, Highlands of Tanzania: African Journal of Agronomy, v. 5, south-western United States, in G. Richter-Bernberg, ed., p. 325-332.

Page 18 www.saltworkconsultants.com

Peryt, T. M., C. Pierre, and S. P. Gryniv, 1998, Origin of Warren, J. K., 2016, Evaporites: A compendium (ISBN polyhalite deposits in the Zechstein (Upper Permian) 978-3-319-13511-3): Berlin, Springer, 1854 p. Zdrada Platform (northern Poland): Sedimentology, v. 45, p. 565-578. Peryt, T. M., H. Tomassi-Morawiec, G. Czapowski, S. P. Hryniv, J. J. Pueyo, C. J. Eastoe, and S. Vovnyuk, 2005, Polyhalite occurrence in the Werra (Zechstein, Upper Permian) Peribaltic Basin of Poland and Russia: evaporite facies constraints: Carbonates and Evaporites, v. 20, p. 182- 194. Pierre, C., 1983, Polyhalite replacement after gypsum at Ojo de Liebre Lagoon (Baja California, Mexico); an ear- ly diagenesis by mixing of marine brines and continental waters: Sixth international symposium on salt, v. 1, p. 257- 265. Salvany, J. M., J. Garcia-Veigas, and F. Orti, 2007, Glau- berite-halite association of the Zaragoza Gypsum Forma- tion (Lower Miocene, Ebro Basin, NE Spain): Sedimen- tology, v. 54, p. 443-467. Salvany, J. M., and F. Orti, 1994, Miocene glauberite de- posits of Alcanadre, Ebro Basin, Spain: sedimentary and diagenetic processes, in R. W. Renaut, and W. M. Last, eds., Sedimentology and geochemistry of modern and an- cient saline lakes, v. 50, SEPM/Society for Sedimentary Geology Special Publication, p. 203-215. Schaller, W. T., and E. P. Henderson, 1932, Mineralogy of drill cores from the potash field of New Mexico and Texas, Bull. U.S. Geol. Surv., No. 833, 1-124, 39 pis., Washington, D.C., p. 171. Smith, D. B., 1974, The origin of the Permian Middle and Upper Potash deposits of Yorkshire; an alternative hypoth- esis [with discussion]: Yorkshire Geol. Soc., Proc., v. 39. Smith, D. B., G. M. Harwood, J. Pattison, and T. H. Pet- tigrew, 1986, A revised nomenclature for Upper Permian strata in eastern England, in G. M. Harwood, and D. B. Smith, eds., The English Zechstein and related topics, Geological Society of London Special Publication no 22. Smith, F. W., J. P. L. Dearlove, S. J. Kemp, C. P. Bell, C. J. Milne, and T. L. Pottas, 2014, Potash – Recent explora- tion developments in North Yorkshire, in E. Hnger, T. J. Brown, and G. Lucas, eds., Proceedings of the 17th Ex- tractive Industry Geology Conference, EIG Conferences Ltd.( 202pp), p. 45-50. Stewart, F. H., 1949, The petrology of the evaporites of the Eskdale no. 2 boring, east Yorkshire; part 1, The lower evaporite bed: Miner. Mag., v. 28, p. 621-675. Stewart, F. H., 1963, The Permian lower evaporites of For- don in Yorkshire: Proceedings of the Yorkshire Geological Society,, v. 34, p. 1-44.

Page 19 www.saltworkconsultants.com

Saltworks Consultants Pty Ltd ABN 068 889 127 Kingston Park, 5049 South Australia

Email: [email protected]

Web Page: www.saltworkconsultants.com

Page 20