Camps, catering and the community joint venture of the Salta puna
It is some eighteen hundred people sleeping every night in construction camps in the Salta puna, and that is the second largest of this province's nine markets. It is also the most misread, for two reasons. The first is that there are two different businesses inside it and one switches off: the construction camp empties when construction ends, on a date written into the resolutions, while the operations camp is small and perpetual. The second is geographic and changes the product entirely: one of the province's plants is not in the puna but in the lowlands, a short drive from the capital, so what is sold there is not beds but a canteen — copying the puna method to the lowlands inflates the number and offers a service nobody buys. And what pinches has to be said: the captive share is among the highest this observatory has measured for a Salta service, because the construction contractor brings the site compound inside its fixed price and the operators' own chains resolve part of it without tendering anything in Salta.
The number is bed-nights and meals served at a price per person per day: bed, full board, cleaning, laundry and the camp's waste and effluent. It does not include the accommodation module, which is construction capex and a different set of books. It is built in three blocks that behave very differently — a large one that switches off on a date, a small one that never switches off, and one that does not even sell beds.
None of the three creates this niche's demand — the people working up there create it — but they define who can bill it and how much is left after tax, which in a single-digit margin business is the difference between a contract and a problem. The ones below open in the reforms panel on the home page, with their status and primary source.
enablesSalta: 70/60 local mining procurementIt is the one that defines who counts as a local supplier and the one most misread in this line of business. It sets a preference of 70% of the annual amount contracted and 60% of the payroll, but the section says shall preferentially contract and the law has no penalties section; another section, moreover, lowers that floor to a negotiated range when the environmental study is approved. The real engine of Salta's local content is a different one and it does have teeth: the reward of offsetting infrastructure works against up to half of the quarterly royalty. And the door for the outsider is written in: a joint venture with a Salta partner from 30%.see the reform →enablesSalta cuts the rate 20% for retail and hospitality, and exempts newly registered taxpayers for 12 monthsIt is the one that finances year one and it has a date: a new taxpayer registering voluntarily pays a zero rate for up to twelve months, and it expires at the end of 2026 unless extended. In none of the nine niches does it matter as much as in this one, and the reason is arithmetic: catering works on a single-digit net margin, so the mining support services rate — with no permanent exemption — takes between a third and a half of the bottom line, and there is no cost of goods to dilute it because it taxes revenue and not value added. Twelve months of grace are, literally, the year in which the business gets built.see the reform →touchesSalta ratified first, and its Gazette publishes the annex that Catamarca’s does not: the 50/50 split is there in writingIt creates no demand for beds or meals, but it defines the perimeter of part of the number: it orders the area Salta shares with Catamarca, where the brine field of one of the projects underpinning the construction block sits. Careful with the ninth clause: the agreement ceases to have effect once Congress settles the boundary between the two provinces, and which side the site ends up on defines which suppliers' roll applies. In its favour, the protocols are declared a reference base for future projects in the border area.see the reform →This market does not float on its own: concrete megaprojects drive it. These are the ones moving demand for this niche — each with its investment and status.
The largest RIGI commitment in Argentine lithium: Rio Tinto is building a 53,000 tpa plant at the Salar de Rincón using DLE (nanofiltration)…
see the project →Gold and silver mining in the puna, on the Salta/Catamarca border area: feasibility development, a 3.15 Mt/yr processing plant and full infrastructure…
see the project →Second stage of POSCO’s Sal de Oro lithium complex in the Salar del Hombre Muerto, on the disputed Salta/Catamarca border strip: it adds a 23,000…
see the project →Who splits the market, where you get in, what pays and what could break it.
already in
split
Segment 1 (construction and modular supply), and the observatory's counter-intuitive finding: the incumbent is from Salta, this layer is NOT imported. In mining since 2002. Rio Tinto's camp at Rincón: 26,000 m2 in 18 two-storey buildings, a 1,000 m2 kitchen sized for more than 2,000 people, a 1,000 m2 dining hall, 600 m2 of laundry, 1,200 m2 of recreation, a 700 m2 project office, a 350 m2 fire station, a 1,400 m2 workshop and reagent shed verif the contractor's own project sheet, opened with our own eyes on 2026-08-05, 4.4 «MV» of own generation verif ibid. — the source writes MV, which is not a unit of power: it will be MVA or MW, and it is an error of the source that is flagged and not corrected — at 3,600 m above sea level, in the 2025-2026 window the sheet declares. Track record: La Alumbrera (2002-2008), POSCO Argentina (2019-2020), TSK/GENSUN PV plant at Olacapato (2019-2020), Eramine Sudamericana (2022-2023), Livent/Minera del Altiplano (2022-2023). Note on domicile: our survey places it in Cerrillos; the website declares addresses in the city of Salta (Juan M. Leguizamon 1946) and Jujuy — not contradictory (plant vs office) but declared.
Segment 1. Listed in the CAPEMISA register under 'manufacture of accommodation modules' verif; dry modular construction from shipping containers unconf.
Segment 1. Presents itself as the only firm in northern Argentina dedicated exclusively to modular construction for mining, with two-storey units suitable for puna conditions unconf.
Segment 2. The TERMINATION of its service set off the first Tolar Grande dispute, and AOMA Salta warned it could call a FULL STRIKE at Mansfield prob. It is the proof that in this line of business the contract rotates.
Segment 2. It received the second Tolar Grande dispute in twelve months, over failure to hire local labour prob. Lindero changed food service provider twice in one year: that is the niche's way in AND the life expectancy of whatever contract is won.
Segment 2. Full-service catering for mining companies; declares itself a Local Supplier of the Salta Puna prob.
Segment 2, and it is the joint venture's natural partner, not the competitor. Based in San Antonio de los Cobres, president Gabriela Miranda. Members from Estación Salar de Pocitos, Olacapato, Tolar Grande, Santa Rosa de los Pastos Grandes and San Antonio de los Cobres — the exact towns that block the road when the contract goes to outsiders. Cases named by the local press: GVH, Answer Group, Planeta Puna, Transportes y Servicios Mineros, Waira Huasi (catering, logistics, modules, chemical toilets, health services, nutrition consulting, internet) prob.
Active in the Argentine mining segment, but our survey could NOT verify a single contract in the Salta puna unconf. If one of them already holds the Rincón camp, the captive share is at the 65% ceiling and the SAM falls by ~USD 4 M/year. Declared, not resolved.
CAPPROMIN complained in Feb-2026 that POSCO brings in Korean suppliers and turns them into service companies once construction ends: this is not future competition, it is a competitor already inside and resident. IMPORTANT COUNTER-EXAMPLE: Rio Tinto bought its camp from a Salta firm (Saltapor), which means the in-house chain is NOT an iron law in this line of business.
A cooperation consortium executing SBDF, earthworks and sludge ponds at Rincón verif. An EPC quoting 'all-in' folds the site compound, canteen and transfers into its price: that is the incumbent the entrant actually displaces, not the catering multinational.
The construction camp is closed and it also has an expiry date: it is the largest block, it comes inside the contractor's price and it switches off in 2029, so it is not a future target either. The operations camp is a different thing and it is winnable — it is step 4, because that contract changes hands when it expires. If running camps is your capability, that is your segment, together with step 6: the corridor lodge is the only product in this niche that does not depend on winning a contract from a miner. The rest you get in through what is renewed and through what does not require altitude, in this order:
The lowlands canteen, in General Güemes — the cheapest door in the whole line of business and the most underestimated. It is at 800 metres above sea level, 50 kilometres from the capital, with no camp, no altitude logistics and no high-altitude medical fitness requirement for staff. You sell food, not hospitality, and the industrial park has other companies around it as well.
Housekeeping and industrial laundry as a subcontractor to whoever holds the contract today — you get in without disputing the main contract, which is the fight an entrant does not win. It is the step that shows you the site from the inside, builds the track record and does not require tying up capital in modules.
Camp waste and effluent management — the line almost nobody looks at, with growing demand from regulatory and lender pressure. It sits right alongside the previous one and is sold to the same buyer, so the cost of sale is already paid.
Full operation of a site camp — the high ticket and the two-to-three-year target. Here the contract changes hands when it expires, which is exactly the opposite of the construction camp: you win it on renewal, not on construction. And there is an actionable, zero-cost data point, which is the most valuable thing on this page: the province's largest camp is already built and has on the order of 700 people sleeping there today, with a declared peak of 2,500 to 3,500 — but who operates that camp today does not appear in any public source. That is the most expensive gap this analysis leaves and the first call an entrant would make: you do not have to wait for an investment decision to find out when that contract expires.
The joint venture, and there are TWO distinct routes worth not confusing — the shortcut the local content law itself leaves open, with a Salta partner from 30%. For a national catering operator that does not want to move its company to the province, that is the way, and it is already proven at the salar project in another line of business. The first route is to partner with the current holder of the contract, by subcontract or consortium. The second — and the partner is already identified, with a name and a geography — is the Puna suppliers' chamber: around 70 companies based in San Antonio de los Cobres and in the towns that block the road when the contract goes to outsiders. They are not the competitor, they are the pool of partners; and they are also the ones who bring the local payroll section 18 requires. Whoever builds that structure once replicates it across five salars.
The corridor's 40-to-80-bed work lodge — the lowest-capital product in the whole niche, and the only one that does not depend on winning a contract with a mining company. Between Tolar Grande's 40 beds and the 22,653 hotel beds of the city of Salta there is practically nothing, and there is demand with nowhere to sleep today: auditors, vendor inspectors, technicians passing through, staff on temporary assignment, a drill rig's second shift. None of them gets into the operator's camp and none can do 5 to 7 hours of road to get back. San Antonio de los Cobres and Estación Salar de Pocitos are the two points, and the Puna gas pipeline already runs along the corridor — Pocitos, Olacapato and Tolar Grande — which is what makes heating viable without trucking in diesel.
~USD 37-44 M/year (55-65% of the TAM midpoint), with four documented capture sources: (a) the operator's own chain — POSCO with Korean suppliers it turns into service companies once construction ends (CAPPROMIN complaint, Feb-2026), Ganfeng with a Chinese chain, Eramet with an in-house structure: this is the portion of Sal de Oro and Mariana that is not tendered in Salta; (b) the mining catering multinationals (Aramark, Sodexo, Compass/Eurest, Newrest), active in the Argentine segment but with no verified contract in the Salta puna — which is why the captive share is declared as a wide band and not as a number: if a global player is already at Rincón, the captive share sits at the 65% ceiling; (c) what the construction contractor brings inside its 'all-in' price (site compound, canteen, on-site transfers) — Milicic + AGV at Rincón is the live case; (d) what the operator internalises (HS&E, medical service, site security), which is already excluded from the TAM but it is worth saying that not the whole camp is outsourced.
~USD 24-31 M/year (35-45% of the TAM midpoint), midpoint ~USD 27 M/year, addressable by a local or national entrant: camp operation, catering, housekeeping, industrial laundry, camp waste and effluent management, and maintenance of the modules. It is equivalent to ~1,370 BED-NIGHTS PER DAY served across the whole province and among all bidders put together. BEWARE THE 70%, THE 60% AND THE 21.02%: the local content of Act 8164 (70% of the amount contracted / 60% of the payroll) and the Sal de Oro II supplier plan (21.02% of the amount for suppliers, goods and works, against the 20% floor of the national RIGI) enlarge the SAM and are a real, auditable sales argument — they are in the text of the resolution — but they are commitments on the investment amount, not on this line of business. They are not a quota, they are not guaranteed demand, and the operator can meet the 70% by buying earthworks and transport, which is where the volume is. Confusing them with own demand is the classic error of this calculation.
USD 3-7 M/year for ONE entrant within 2-3 years = 150-350 bed-nights per day. Realistic composition: (year 1) a catering contract for a mid-sized camp, or the housekeeping + laundry package of a large one, entering as a subcontractor to the current holder or through a joint venture; (year 1-2) the lowlands canteen in General Güemes, the lowest physical barrier; (year 2-3) full operation of a 300-600 bed camp under its own contract. It is a 40-90 person SME, it is replicable to Jujuy and Catamarca, and it is not a unicorn.
When you get paid, and what blocks it
effect
It is the niche with the highest employment per dollar billed in the entire Salta mining chain, and by a distance. A full-service camp employs on the order of 1 worker per 5-8 residents estim derived from the Diablillos table 21-15 itself: camp personnel USD 31.2 M LOM ÷ 14 years = 2.23 M/year against a resident population of the order of 400, with a declared assumption of an annual labour cost per worker of USD 28-45 k — without it the ratio does not hold. Over an addressable market of ~USD 27 M/year (≈1,230 residents served) that gives 155-250 jobs in the province, against the dozens left by an engineering niche. This is the niche where the neighbour from Tolar Grande gets into mining.Concrete trades, with a route and without a university degree: camp cook, kitchen assistant, baker, dining-hall waiter, housekeeper, industrial laundry operator, operator of the camp's water treatment and effluent plant (the camp's water permit is a procedure of its own before Water Resources with publication in the Official Gazette: the Mariana file we opened is exactly that), module maintenance (electrician, plumber, gas fitter, REFRIGERATION TECHNICIAN), storeroom and warehouse supervisor with cold chain, internal transfer driver, camp manager. And two from the degree band, which is precisely where the ILO detects Salta's shortfall: DIETITIAN and food safety technician (HACCP/food hygiene) — because the ILO diagnosis says Salta has an oversupply of professionals and managers and a shortfall in health verif.Training: the gap is named and not executed. UPATECO has 70 training offerings and 2,300+ enrolments in 2026, but no campus in the puna (city of Salta, Rosario de Lerma, Rosario de la Frontera) prob. Tolar Grande certified 10 assistant electricians — ten people, in the town that supplies labour to Taca Taca, Mariana and Lindero. And the ILO left in writing the recommendation nobody executed: 'to assess the creation of a puna training centre' verif. A camp operator that trains in-house is not doing CSR: it is manufacturing the payroll that Act 8164 requires of it (the 60% of section 18, the 80% of section 16).Local linkage, and it is the widest in the chain: catering buys food, which means it is the only mining link that connects directly with the Salta valley (growers, bakeries, meat processors, refrigerated transport) — with the NON-mining economy. And community management capability is proven at small scale: the Mariana project's shop is run and administered by the Tolar Grande community prob.What is not resolved, said head-on:(1) the community wants the contract and today cannot take it. Tolar Grande has a little over 200 inhabitants and a 2,000-bed camp needs 250-400 workers: the whole town is not enough. The 30% joint venture solves OWNERSHIP, not the payroll; selling the opposite is exactly what produced the two disputes.(2) It is the chain's most massive employment and also its worst paid: in the only company-level collective agreement with Salta scope (CCT 1614/19 'E'), general services is category A, the lowest on the scale verif; the average Salta mining wage of ARS 6,586,941 (Mar-2026) is for the industry, not for the camp kitchen, and saying 'mining job' without saying the category is inflating.(3) It is the most cyclical niche in the Salta observatory: 61% of the TAM switches off in 2029 and there is no backlog — an empty bed is not billed the following month, and Rincón Lithium suspended is the live demonstration.(4) Contract turnover cuts both ways: Lindero changed provider twice in twelve months, and that is the way in AND the life expectancy of whatever contract is won.(5) The margin is not measured: there is no public unit price, there are no public payment terms and the turnover tax classification of the line of business is not settled — the market can be sized, the business can NOT yet be sized, and that is the difference between this document and an investment plan.
calculate it
Concentration Two layers with opposite structures, and the entrant picks the wrong layer if it does not separate them.Segment 1 - construction and modular supply of the camp: concentrated, and with a fairly uncommon favourable anomaly — the incumbent is from Salta. There is no public register of awards, so an HHI would be invented; what can be stated is that, of the large documented camp projects in the Salta Puna, those with a known builder have THE same builder (Saltapor, with Rio Tinto, POSCO, Eramine, TSK/GENSUN, Livent and La Alumbrera on its record). Local construction capability exists and is documented with a name and with completed work. Behind it there are two more local bidders (AltoAndina, Mari Mari Modular) with no named contract.Segment 2 - camp operation and catering: fragmented, rotating and with no stable incumbent, with hard evidence of turnover — Lindero changed food service provider twice in twelve months (Cookins -> Central Andino) and no Salta operator has a published catering incumbent. On the supply side there are ~70 CAPROSEMITP members atomised across the Puna towns, plus Catering Andino, with none holding a named full-service contract. That is what makes the niche enterable — and also what makes the contract won fragile.Segment 2's final share is NOT split by the visible bidders: it is decided by two things that are not public data today — (1) whether any mining catering multinational (Aramark/Sodexo/Compass/Newrest) is already inside the Salta puna unconf declared gap, and (2) how much of the canteen and site compound still travels inside the 'all-in' price of an out-of-province construction contractor.Beware the easy reading: catering and cleaning are among the 73 of 91 categories with local supply (our own survey of the supplier ecosystem, against the 18 without supply, which are heavy machinery, SAG mills and crushers). This niche's gap is not one of existence: it is one of scale, qualification and corporate vehicle.
There are four clients with four different pockets, and the order in which you approach them decides whether the business starts fast or dies waiting on the hardest door:
The first invoice and the lowest physical barrier in the whole line of business: at 800 metres above sea level there is no altitude medicine, no camp and no puna logistics. It is the same industrial park where the province's two lithium chemical plants sit, on a national route and with a railway. The volume is not small — one project declared 3,600 contractors at its peak before the provincial Senate — but the unit of that figure is not established, so it was sized with a wide band rather than picking the number that makes it bigger.
You get in without fighting for the main contract, which is the fight an entrant does not win. And there is a concrete reason to prefer this door to the operator's: the provincial mining suppliers' roll has 498 suppliers with an approved application, so the intermediate buyer is already identified and approved. The cost of sale is low because the client already knows what it needs: it lacks capacity, not a supplier.
The high ticket and the two-to-three-year target. Here the advantage is calendar and not price: the contract changes hands when it expires, which is the opposite of the construction camp. And the filter deciding who can bid is the local content regime: actual and tax domicile in the province, 80% of the payroll here, and the joint venture with a Salta partner from 30% as the door for an outsider.
It is the largest block and the hardest, because it is not bought separately: it travels inside the fixed price of the works. You win it by partnering with whoever already has it, not by competing against them. And it has an expiry date written in: it switches off with the legal investment deadlines of the resolutions. Whoever sizes a company counting on this block is left with excess structure in 2029.
It is the only payer in the niche that does not tender: it pays by the night, against an invoice, with no supplier approval and no need to qualify as a local supplier. Auditors, vendor inspectors, technicians passing through and drill rig second shifts have no bed today between Tolar Grande (40 beds) and the capital (22,653). It is the smallest ticket and the one with the least competition, and that is why it is the natural door for local capital that cannot fight for a full-service contract.
This is not «what breaks it»: it is the dashboard for not ending up with excess structure. In this niche the expensive mistake is not coming in late — it is sizing the company on the construction peak and discovering in 2029 that half the market had a date on it.
It is the most honest thermometer in this line of business, because this market is sized by people and not by dollars invested: every person entering the mining roll is demand for a bed or a meal. But it has to be read with two cautions almost nobody applies. First, the figure is registered employment and the site moves twice as many people: the seven-by-seven and fourteen-by-fourteen regimes require two crews, so the roll of people is close to double the number up there at any one time. Second, not all of them sleep at altitude — the part working in the lowlands consumes canteen and not beds, which is a different product and a different price. And the sign has to be stated: that number comes off a 5.7% year-on-year fall, and Salta was the only one of the seven provinces with the largest mining employment to fall in March 2026 — while Catamarca rose 21.9% and San Juan 8.4%. It is still third in the country behind Santa Cruz and San Juan, but whoever sizes against this market has to reckon with a roll that shrank, not one that expanded.
Salta's Directorate General of Statistics and the national registered employment record, published quarterly ↗Four companion signals that move the number without warning, and the first needs nothing to happen: who holds the operations contract for the province's largest camp today. It is live, it was not published, and finding out costs two phone calls — it is the only signal on this dashboard that is resolved by your own work and not by waiting on a third party. The final investment decision on the silver project, expected in 2027: until it happens, its peak of twelve hundred people does not exist, and when it happens the construction block jumps. The arrival of a global mining catering operator at the largest site: today there is no verified contract in the Salta puna, and if one appears the captive share jumps to the ceiling of its band overnight. And the row of the schedule that decides the tax classification of camp catering — it does not change the size of the market, but in a single-digit margin business it decides whether the contract leaves money or not.
The province already has the case: the Rincón Lithium plant (Argosy / Puna Mining), 2,000 t/y, was recorded as suspended as of the Aug-2026 cut-off prob own survey of the province's plants; there is no operator statement or administrative act backing it, against the backdrop of the roughly 80% fall in the lithium price. A suspended camp bills zero the following month — there is no backlog to cushion it, because the unit of sale is the night. It is the fastest reaction to the cycle in the entire services chain.
The DFS declares the full schedule: early works Q3-2026 -> target financing Q4-2026 -> FID Q2-2027 -> construction from 2027 with a 2028-2029 peak -> first production before the end of 2029 prob. Diablillos is ~27% of block A's capex and 100% of its 1,600 planned beds. If the FID slips, block A loses ~USD 8-10 M/year off its midpoint. It is the most measurable killer and it has to be watched quarter by quarter.
The USD 5,250 M and the 4,000-person peak are the niche's biggest upside (+USD 88 M/year at peak, more than all the rest combined) and there is no public evidence that the project has filed its application to the large-investment incentive regime —its owner stated in Feb-2026 that it was preparing to file it, and the official portal does not publish the detail of the projects under evaluation, so non-filing is not verifiable there unconf status of the filing: it is the softest ring in the portfolio. An entrant that sizes capacity against Taca Taca goes bust waiting.
The suppliers' chamber CAPPROMIN reported in Feb-2026 that POSCO brings in Korean suppliers and turns them into service companies once the works are finished prob industry press. If the pattern the complaint describes holds, the competitor is not a future threat but one already operating here, with Argentine residency — and that is exactly what we watch. And Ganfeng solved the province's largest solar park with its own chain: USD 190 M of infrastructure that was never tendered in Salta prob company and business press — that is a precedent in infrastructure, not a measurement of camps or catering. Honest flip side: Rio Tinto bought the province's largest camp from a Salta local, so the captive chain is no iron law in this segment.
The demand for modular construction at Rincón — the largest documented package in the province, 26,000 m2 in 18 buildings — HAS ALREADY been collected. What remains alive in that asset is its operation. An entrant aiming to manufacture modules arrives late for Rincón and early for Diablillos (whose 1,600 beds depend on the Q2-2027 FID).
61% of the TAM is construction and it switches off with the resolutions (Rincón 30-Jun-2029, Sal de Oro II 31-Jul-2029). What remains is the perpetual core of operation: ~USD 24 M/year, half of what falls away. Finding: that core is far bigger than expected (operation ~2x smaller than construction, not ~10x) because Salta already has four lithium plants producing and a gold mine running. Whoever does not build the operations line closes in 2030; whoever does keeps the client.
Jul-2025: Cookins' service terminated at Lindero, with AOMA Salta warning it could call a FULL STRIKE at Mansfield prob. Jan-2026: second conflict at Tolar Grande in twelve months, with protests against Central Andino over the lack of local hiring prob. It cuts both ways: it is the way in for whoever brings local hiring, and it is the risk for whoever already won - a blockade on RP 27 or on the access to Tolar Grande halts camp resupply within 24-48 h, and an unsupplied camp gets evacuated.
In Salta mining extraction pays 0.75% turnover tax (and 0% with an exemption certificate) and mining support services (code 99000) pay 3.60% with no permanent exemption verif the schedule of activities of Salta's Directorate General of Revenue: 4.8 times the mine's rate. In catering the typical net margin is single-digit: 3.60% on gross revenue eats on the order of 35-50% of the bottom line estim of a contract with a 7-10% net margin, and there is NO cost of goods to dilute it because the tax is on revenue, not on value added. And the alternative classification, which was flagged as an open gap, is now open and it cuts against us: if the tax authority treats it as a food service rather than as mining support, it pays MORE, not less. Catering has its own code — 562010, 'preparation of meals for companies and events', which the text defines as including 'catering services' — and pays 4.50%; a canteen inside the establishment (562091) pays the same verif Annex I p. 39. Worse still: those are precisely the two codes that the implementing regulation carves out by hand from the Ley 8496 discount that does reach the rest of hotels and restaurants verif art. 1(c). The classification decides between 3.60% and 4.50%, and it is the first check before quoting.
The Supreme Court assumed original jurisdiction on 17-Oct-2025 in the injunction brought by the Roundtable of the 33 Kolla and Atacama communities (~7,000 people), with an interim suspension of permits pending decision prob. It does not touch the ring A salars today, but it is the only front with the capacity to halt permits in the province.
Bed-nights per year multiplied by a price per person per day, calculated separately for construction, operation and the lowlands because they are three markets with three different clients. The volume was built along three routes — capex by intensity, declared headcounts and installed physical capacity — which fall in the same band without being independent of one another: they share the two projects that feed them, so their agreement counts as order of magnitude and not as cross-validation. The price is the link with no source and it is declared as such.
A robustness check along three paths. This niche arrives with one strength and one debt: the volume is already built along three in-house routes —which fall in the same band without being independent of one another, and that is why the construction block was corrected 14% downwards—, and the price is —in the calculation's own words— the only link in the whole chain without a source. That is why all three paths aim there: not at how many people sleep up there, but at what the night costs. (a) The two twins that computed the same price by another route, and do not agree. The assumption here is USD 45-75 per person per day (midpoint 60) for bed plus full board, cleaning, laundry and effluents. San Juan builds it from the bottom up —3.0 to 3.2 meals at USD 8-12 plus USD 8-15 of hotel services— and publishes USD 35-55. Río Negro builds it the other way round, separating lodging from food: USD 35-50 of lodging plus facility management, food excluded, and USD 12-30 of catering, that is USD 47-80 combined. What matters is the effect on the headline, and it reads without touching the volume: swapping the price alone, the Río Negro midpoint —USD 63.5— moves the TAM to ~USD 72 M/year and the San Juan one —USD 45— moves it to ~USD 51 M/year. Both land inside the published 51-86 band, and the San Juan one lands exactly on its floor. When two twins disagree the reason has to be stated, and here it can be: the difference is one of scope, not of market — the USD 8-15 San Juan calls “hotel services” is housekeeping and does not cover the bed, which is exactly what Río Negro charges separately at USD 35-50. On that reading, the comparable twin is the Río Negro one and it validates the level; the San Juan one measures something else and is declared as such. (b) The external anchor, and it is the hardest this sector has anywhere in the repository. It is neither a benchmark nor an assumption: these are purchases actually executed. VMOS reported $4,059 M on packed meals in the first quarter of 2026 across ~2,550 direct workers. Annualised at the stated exchange rate of ~$1,500, that is USD 10.8 M over 930,750 person-days, or USD 11.6 per person per day of food. The lowland block here —General Güemes, canteen without beds— uses USD 8-14, midpoint 11: the only unit service price in the repository with a real purchase behind it lands practically on the assumed midpoint. And the scopes are deliberately comparable: it is tested against the lowland and not against the puna, because VMOS operates at sea level and Güemes sits at 800 metres, 50 km from the provincial capital. What this path does NOT validate is the high-altitude price, which governs most of the figure: feeding people at 4,000 metres with Sico-pass logistics does not cost what feeding them at an Atlantic port does, and that gap remains unmeasured. (c) Weight within the province: it does not discriminate, and is declared as such. This market weighs 29.7% of the USD 229.3 M/year of Salta's nine investor markets; the San Juan pair —camps plus catering— weighs 19.8%. The gap says nothing about the niche because the denominators are not comparable: San Juan's is almost five times larger and dominated by power and road works, two sectors that in Salta do not yet have the scale of San Juan's copper. A share taken over baskets of different composition is not a control. What still stands: the volume rests on three in-house routes already corrected for their overlap —and it is the province's second-narrowest band in relative terms, behind Puna logistics—; the lowland price is confirmed by a real purchase; and the high-altitude price is supported by a comparable twin, not verified. The gap is concrete and has the shape of a single document: one invoice or one tender for a puna camp bed-night. It is the same one the calculation already declares, and it remains the only thing standing between this figure and being anchored end to end.
The number rests on a few variables. The formula shows how it moves when each one changes; and each variable carries its freshness seal — how often it is worth revisiting. estim
Every figure is checked against its source before we publish it. Here we show what backs it — and where the verified data ends and our estimate begins.
The volume does not come from an assumption: it was built along three routes, and then how much their agreement was worth was audited. The first divides capex by an accommodation intensity measured in a primary source — the feasibility study of one of the projects publishes the total man-hours of construction and its regime of fifteen days on and fifteen off, and from there comes how many person-days of presence each million dollars invested generates. The second starts from the concurrent headcounts each site declares. The third starts from the installed physical capacity. The three fall in the same band, but they are not independent of one another, and that is why the volume was corrected downwards: all three are fed by the same two projects, and the physical-capacity route weighted one of them across the whole window when its investment decision only lands halfway through — two figures from the same study that differ by 2.7 times. With that weighting redone, the construction volume falls by 14% and it is the corrected one that is published. The agreement of the three counts as an order-of-magnitude check, not an independent validation. The assumption that governs the number is declared and not disguised: the price per bed-night does not exist published in Argentina. An explicit band was used, and it was cross-checked against the only open breakdown of operating cost that exists for a puna site, which is in the technical report of a project in the province: inverting that calculation, the assumption is covered by the primary source itself. Two precisions that narrow downwards, not upwards. The project sheet of the largest camp declares a kitchen and dining hall sized for more than two thousand people per day, and that is meal-service capacity — it does *not* say beds or how many people sleep there, so the headcount was taken as our reading and not as a figure from the source. And the five thousand jobs figure that circulates for one of the plants was not used: with it the lowlands block would come out inflated by almost 40%. What was deliberately left out, so as not to charge twice for the same service: personnel travel and on-site transport — which are counted in this same province's logistics market — plus the medical service and health and safety, which have their own niche. If they were added, the construction block would rise by around 40%.
How to cite this figure: Despegue (2026). Camps, catering and the community joint venture of the Salta puna · Salta. despegueargentina.com/en/salta/campamentos-catering-mineria-altura · terms of use
- 10+ years in data science for clients across Europe and the Americas
- Certified in AI governance (ISO/IEC 42001)
- Machine Learning (Google Cloud)
- Registered expert with the European Commission
Get on board the takeoff
This week’s updates: the map of camps, catering and the community joint venture of the Salta puna and the niches opening up, related courses and new provinces as they launch. Free.