Could a company pool Java’s small farm plots into land blocks for PLN’s solar plants?

Written by

in

A solar power plant needs land: a large, unbroken block next to a substation where its power can enter the grid. In October 2026 PLN, Indonesia’s state electricity company, which buys the power such plants produce, published how much of that ground its 100-gigawatt solar programme needs: 210,474 hectares at 1,015 substation points, of which 107,086 hectares sit at 485 points in Java, Madura and Bali [10]. The programme has drawn foreign interest, with investment estimates above Rp1,140 trillion [1].

PLN does not buy that land. It opened a register, the DPT Tanah (a list of pre-qualified land providers), where owners can offer their ground for lease, or as a share in the project, to whichever developer PLN later selects [10]. By 21 October, 31 parties had signed up and nine of them had offered 4,557 hectares between them, about 2% of the national need. Only two of the 31 were individuals [10]. One rule explains part of that: an individual may register only with at least one hectare in a single stretch [10]. Indonesia’s 2023 farm census counted 16.89 million farming households working less than half a hectare [13], and we infer that much of the farmland around Java’s substations is held in plots of that size.

Who owns the ground next to a substation

Three kinds of party meet at each point. PLN decides where plants go and buys their electricity under a long-term power purchase agreement. A developer, either a private independent power producer or a PLN group company, wins the right to build the plant in a PLN tender, largely on the lowest tariff it bids [9]. And the landowners own what the plant will stand on. Under PLN’s procurement, the developer is expected to secure all the land, for the plant and for the line to the nearest substation [2]. Land feeds straight into the electricity price: PLN’s chief executive told parliament that land at Rp200,000 per square metre adds about 1 US cent to each kilowatt-hour, and Rp600,000 adds about 3 cents [6].

That is why PLN prefers free state land. But of 28,000 hectares the land ministry proposed in Java, PLN found about 8,500 technically suited to its grid [6], set against the 107,086 hectares its Java points need [10]. Most of the remainder will have to come from private owners.

Land for PLN’s solar points in Java against what is on offer

Java-Madura-Bali need107,086 ha
State land proposed in Java28,000 ha
State land technically fit8,500 ha
Offered in register, all Indonesia (21 Oct)4,557 ha
Java-Madura-Bali need, state land proposed and found suitable, and the hectares offered nationally in the register’s first weeks.Sources: [6] [10].

Why the land arrives last

The mechanism, in one sentence: PLN picks developers on price before anyone has secured the land, while the land at Java’s points is split among hundreds of owners too small to register on their own, so a contiguous block gets assembled late, under deadline, one parcel at a time.

The cost of that order shows up after the award. The Indonesian Solar Energy Association (AESI), the developers’ trade body, said in August 2026 that projects “awarded for construction based on lowest tariff bid” had “subsequently fell through because land could not be acquired” [9]. Earlier analysis for the Energy Transition Partnership, a donor-funded programme, found that two-thirds of utility-scale solar projects were already running one to three years late [4].

Today developers deal with this themselves. Their own land teams and local brokers negotiate parcel by parcel once a site is theirs, or they hire a consultant. PT Global Solusindo Gemilang (GSG), a Jakarta firm, advertises land scouting, measurement, owner coordination and permits for power and solar developers, and names work on a West Java gas plant [3]; it publishes no prices. Engineering firms such as PLN Enjiniring, PLN’s own engineering arm, and Applus+ survey sites and grid connections, but do not tie up owners [11][12]. Large holders, such as the coal miner PTBA with its reclaimed mining land, can offer whole blocks directly [14]. All of these start after a developer is engaged, or serve owners who already hold a block. Nobody in the sources pools the half-hectare owners before the tender.

One substation point, assembled in advance

The business we examine, Ready-to-Lease Village Land Blocks, would do that work first and sell the result to whichever developer wins. Take one Java point of average size, about 221 hectares (107,086 hectares over 485 points [10]). With plots of around half a hectare, that is several hundred owners; we assume 300 to 600.

How one point would be assembled before the tender

  1. 1. Pick a point
    A PLN-mapped Java point without state land, screened on spatial and tax maps
  2. 2. Organise owners
    Village meetings; several hundred owners join one vehicle they own
  3. 3. Verify
    One drone map, neighbour boundary sign-offs, risk-scaled title checks
  4. 4. Sign
    Standard leases or share contributions, valid only if a developer is selected
  5. 5. Register
    The block enters PLN’s DPT Tanah list as one contiguous holding
  6. 6. Sell
    The selected developer signs one lease and pays the company per hectare
  7. 7. Pay rent
    Owners receive rent or shares through their vehicle for the lease’s life
The steps the proposed company would take at an average 221-hectare Java point, and when money changes hands.Based on sources [10] [2] and our own reasoning.

The company’s field organisers would screen PLN’s point list against spatial plans and tax maps to find points without state land, then hold meetings with owners, convened by village heads. Owners who agree would join one vehicle they own together, a cooperative or a small company, so that the block, not each plot, meets the one-hectare rule. A single drone map of the whole area, boundary sign-offs between neighbours and title checks scaled to each parcel’s risk would replace hundreds of separate surveys. Each owner would sign one of a few standard leases or share contributions, valid only if the point is tendered and a developer selected. The vehicle would register the block in DPT Tanah.

Related Business Opportunity:  What if fuel stations in Indonesia could spot delivery trouble earlier?

When PLN tenders the point, the winning developer signs one lease with one counterparty for a verified, unbroken block, and can add the strip for the line to the substation. At signing it pays the company a fee per hectare; owners receive rent, or project shares, through their vehicle for the life of the lease. Solar power contracts in Indonesia run 20 to 25 years [9]. What the developer gets that it does not get today is tenure before its tariff and financing deadlines start running. What the owners get is access to programme rent without selling and without paying the document costs PLN leaves to registrants [10].

What the fee has to cover

No record of what developers pay for land work today could be found: not GSG’s fees, not a tender, not a disclosed contract. So the price here is our assumption: Rp15 million to Rp40 million a hectare, a small fraction of the Rp2 billion to Rp6 billion a hectare that PLN’s own land-price examples imply. At an average point that is about Rp3.3 billion to Rp8.8 billion (roughly US$200,000 to US$530,000). On our assumptions the work itself, organisers, mapping, legal checks and agreements, costs about Rp4 million to Rp8 million a hectare.

The relationship that decides the business is not the margin on a paid hectare but how many prepared points ever pay. On our assumption the company spends six to nine months organising a point before PLN tenders it, and registration does not guarantee the land will be used [10]. If only half of the points it prepares are tendered and won by a developer who pays, the cost per paid hectare doubles to about Rp8 million to Rp16 million. At the low price, the business breaks even only if roughly a third to a half of prepared points pay.

The first market is countable. If PLN’s 485 Java-Madura-Bali points come to tender evenly over 2027 to 2029, about 160 a year reach selection. A company that prepared and sold two to five average points a year at the assumed fees would earn about Rp6.6 billion to Rp44 billion a year. Near the middle, around Rp20 billion, that funds a regional team of 20 to 40 field staff; at the low end it is a side business. Beyond Java, the same work applies at PLN’s 67,724 hectares of Sumatra points and its battery sites at the same substations, though in Kalimantan a provincial power company, KLKT, has signalled it wants to supply land at that region’s 103 points [10].

Where it could go wrong

The sharpest risk is PLN itself. At the October briefing, KLKT’s representative praised the register for cutting out intermediaries and called for layered checks of every registrant’s identity, title and boundaries [10]. A paid organiser behind an owners’ cooperative could look like exactly the middleman PLN meant to remove. The first registration round closed on 29 October 2026 [10], and no later round is documented.

The second is a change in the tender. AESI has proposed that PLN take developers’ plans first and give them up to a year to secure land before the final tariff is set [9]. If PLN adopts that, consultants such as GSG could do the same pooling after a plan is accepted, and the advantage of starting early narrows. Rivals could also copy the method: what the company would hold is signed conditional leases at specific points and trust in specific villages, which matter at those points but not at the next one. Over time, a track record PLN accepts and the steady income of administering rent for decades could make each next point cheaper, but that is a path, not a given.

The third is the owners. Several hundred households must agree to tie up farmland for two decades at a rent nobody has yet published, and a few refusals in the middle of a block break it.

So could a company pool Java’s small plots into blocks for PLN’s solar plants? The need is documented and nobody yet does the pooling before the tender, so the opening is real. Whether it is a business depends on two answers anyone can get cheaply: a written question to PLN’s DPT Tanah team on whether a pooled owners’ vehicle may register, and in which future rounds, and a priced offer for one mapped West Java point sent to three to five developers already on PLN’s list. If PLN says no, or no developer will pay more than about Rp15 million a hectare, the idea fails.

Related Business Opportunity:  Could a company get Riau's independent oil-palm growers the price mills pay their partners?

What is known, and what is not

Documented: the land PLN needs, the one-hectare rule, the shortfall of fit state land, that land stays the developer’s job, and that awarded projects have failed on it. Inferred: that most Java points sit on fragmented private plots, and that a developer would value tenure secured before award. Unknown and decisive: what developers would pay, whether PLN accepts the vehicle, how many prepared points get tendered, and whether owners sign. We would like to hear from former developer land teams, agrarian lawyers and surveyors, and rural cooperative organisers in West and Central Java who can answer them.

What has to be true

The business described above stands on these claims, listed in the order we would test them.

  1. PLN’s DPT Tanah will accept a cooperative or owner-owned company that pools many sub-1 ha parcels, organised by a paid service, and will open registration again after the October 2026 round
    Not yet tested · PLN (DPT Tanah team)
    Business entities may register (29 of 31 registrants) [10]; PLN frames the register as cutting out intermediaries [10]; the first round closed on 29 October [10] [10]
    Would break if: PLN says pooled owner vehicles or organiser-led registrations are not accepted, or no further round opens
    Cheapest test: Written question to PLN’s DPT Tanah team about pooled owner vehicles and future rounds
  2. Selected developers will pay a third party about Rp15-40 million per hectare for a contiguous, verified block handed over before award
    Not yet tested · Developers
    No record of what developers pay for land work today was found; awarded projects have failed on land [9] [9]
    Would break if: Three to five IPPs on PLN’s list decline a priced offer or name a price below the delivery cost
    Cheapest test: Priced written offer for one mapped point to 3-5 DPT-listed IPPs
  3. At least roughly a third to a half of the points the company prepares are tendered and won by a developer who pays
    Not yet tested · PLN and developers
    Registration does not guarantee use [10]; two-thirds of utility solar was already delayed 1-3 years [4] [10] [4]
    Would break if: PLN’s tender packages skip most fragmented private-land points or slip beyond 2029
    Cheapest test: Compare PLN’s first tender packages against the DPT point list to see what share of private-land points is tendered
  4. Most owners at a 221 ha point (300-600 households) will sign a conditional 20-30-year lease or inbreng, with few holdouts breaking contiguity
    Not yet tested · Smallholder landowners
    Lease and inbreng let owners keep title [10]; no owner rent level is published [10]
    Would break if: More than a few parcels in the middle of a block refuse, or owners demand rents developers reject
    Cheapest test: Village meetings at one mapped West Java point with a standard lease offer, counting signed letters of intent
  5. Standard lease plans, one aerial map and risk-tiered checks bring direct cost to about Rp4-8 million/ha while satisfying PLN’s field and document verification
    Not yet tested · PLN and ATR/BPN land offices
    PLN’s verification covers documents and field checks [10]; the cost inputs are assumptions [10]
    Would break if: PLN or BPN require full per-parcel measurement and bespoke agreements, pushing cost above about Rp15 million/ha
    Cheapest test: Price one point’s verification with a licensed surveyor and notary and ask PLN which checks it requires
  6. Most Java points lack technically fit state land and sit on fragmented private parcels
    Our inference from the sources · PLN and ATR/BPN
    8,500 ha of state land is technically fit against 107,086 ha needed in Java [6][10]; 16.89 million farm households hold under 0.5 ha [13] [6] [10] [13]
    Would break if: The state assigns far more fit land or the Java points prove to sit on large holdings
    Cheapest test: Overlay PLN’s Java point list on land-office ownership maps for one district
  7. The tender keeps land on the developer before tariff commitments, so tenure before award stays valuable
    Shown by our sources · PLN and developers
    Developers acquire plant and line land [2]; AESI proposes a year to secure land before the final tariff [9] [2] [9]
    Would break if: PLN adopts a land-first window, letting consultants do the same work after a plan is accepted
    Cheapest test: Read the next PLN tender package’s land clause

Research Notes

The sources, calculations and evidence records behind this investigation. Numbered citations in the article link here; ↩ returns to the text.

Sources

[1] Proyek PLTS 100 GW RI Dilirik China, Investasinya Tembus Rp 1.140 T (detikFinance, 11 Nov 2026) ↩

Read in full. Supports: The programme’s scale and investment interest; PLN’s estimate of about 35 GWp developable to 2029.

[2] JETP Indonesia: Procurement of RE — main issues in PLN’s procurement process ↩

Read in full. Supports: Developers are expected to acquire all land for the plant and its line to the nearest substation.

[3] PT Global Solusindo Gemilang (GSG): land acquisition and permitting consultant ↩

Read in full. Seller’s own page: figures are advertised. Supports: GSG’s advertised land and permit services and named projects (seller’s own page; no prices published).

[4] ETP, 1 GW Solar Mapping and Development Plan: Solar PV procurement challenge / impediments (2025) ↩

Read in part. Supports: Utility-scale solar delays (two-thirds delayed one to three years, citing IESR 2024).

[5] Kontan, 4 May 2026: PLN launches Giga One PLTS 1.22 GW — tender schedule and targets

Read in full. Supports: PLN’s Giga One 1.22 GW tender schedule, an early package of the programme.

[6] Warta Ekonomi, 2 Jul 2026: Garap PLTS 100 GW, PLN wanti-wanti harga tanah ↩

Read in full. Supports: PLN chief executive’s land-price sensitivity (Rp200,000/m² ≈ 1 US cent/kWh) and the 8,500 ha of 28,000 ha proposed Java state land found technically fit.

[7] ANTARA, 29 May 2026: Govt allocates 24,000 hectares in Java for 100 GW solar

Read in full. Supports: The government’s allocation of about 24,000 ha in Java for the programme.

[8] Kontan, 20 Oct 2026: PLN matangkan pengadaan PLTS 100 GWp, tender bertahap

Read in part. Supports: PLN’s staged tenders, including a Bank Tanah project in Purwakarta among six projects of 4.55 GWp.

[9] Argus, 20 Aug 2026: Indonesia solar groups seek better land, permit rules ↩

Read in full. Supports: AESI: awarded projects that fell through on land; the proposed year to secure land before the final tariff; 20–25-year power contracts.

[10] Analisnews, 23 Oct 2026: PLN market sounding DPT Tanah PLTS 100 GWp & BESS, target 210 ribu hektare ↩

Read in full. Supports: DPT Tanah: 210,474 ha at 1,015 points; 107,085.6 ha at 485 Java-Madura-Bali points; the 1 ha rule; lease or share, not purchase; document costs on registrants; 31 registrants and 4,557 ha offered by 21 Oct 2026; round closing 29 Oct; registration does not guarantee use; KLKT’s statement.

[11] Listrik Indonesia, 19 May 2026: PLN Enjiniring siap kawal PLTS Mentari Nusantara I ↩

Read in part. Supports: PLN Enjiniring’s site-potential and interconnection survey role.

[12] Applus+ Indonesia: renewable energy services ↩

Search summary only. Seller’s own page: figures are advertised. Supports: Applus+ renewable-energy services (seller summary).

[13] Bisnis, 4 Dec 2023: BPS (Sensus Pertanian 2023) — petani gurem 16.89 juta ↩

Search summary only. Supports: BPS 2023 farm census: 16.89 million smallholder (petani gurem) households under 0.5 ha.

[14] Kontan, 2026: PTBA has 250 ha of post-mining land, partners Pertamina NRE for PLTS ↩

Search summary only. Supports: PTBA’s 250 ha of post-mining land offered for solar with Pertamina NRE.

[15] Kompas, 27 Aug 2026: PLN IP develops 490 MWp PLTS; 14 projects ~5.3 GW in 6 provinces launched

Read in part. Supports: PLN IP’s 490 MWp of solar and 14 launched projects of about 5.3 GW in six provinces.

Who already does part of the job

Alternative What it offers Price In Java points today? Source
Developers’ own land teams and brokers Parcel-by-parcel negotiation after award Not published Yes, wherever developers build [2][9]
PT Global Solusindo Gemilang (GSG) Land scouting, measurement, owner coordination, permits for developers, hired per engagement Not published Yes: names West Java work [3]
State land (land ministry, Bank Tanah) Free state land assigned to the programme No market price Yes: 8,500 ha fit of 28,000 ha proposed; 24,000 ha allocated; Purwakarta project [6][7][8]
DPT Tanah business registrants (29, unnamed) Land they already hold, for lease or shares Negotiated with developer Unknown: identities and locations unpublished [10]
KLKT (East Kalimantan provincial power company) Clean land provision at Kalimantan points Not published No: Kalimantan focus [10]
PLN Enjiniring, Applus+ Site and grid-connection surveys, permitting studies Not published Yes, for programme sites; no owner tenure [11][12][5]
Large holders (e.g. PTBA) Whole blocks of company land Not published Mainly outside Java [14]
Owners registering alone Free registration from 1 ha contiguous; owner pays documents Free Yes, but 2 individuals by 21 Oct [10]

Searches for an operator already pooling smallholders into pre-tender blocks, in Indonesian and English, found none; that is what was searched, not proof that none exists. PLN group developers such as PLN IP are buyers as well as builders [15].

Economics workings

Step Calculation Status
Average Java point 107,085.6 ha ÷ 485 points ≈ 221 ha Sourced [10]
Owners per point 221 ha ÷ about 0.5 ha per plot ≈ 300 to 600 owners Assumption; plot size from [13]
Land value range Rp 200,000 to 600,000 /m² × 10,000 m²/ha = Rp 2 to 6 billion /ha Sourced examples [6]
Fee Rp 15 to 40 million /ha (0.25 % to 2 % of land value) Assumption: no market fee found
Revenue per point 221 ha × Rp 15 to 40 million = Rp 3.3 to 8.8 billion Illustration
Direct cost Organisers Rp 1.6 m + mapping Rp 0.5 to 1.5 m + legal Rp 1 to 2.5 m + engagement Rp 1 to 2 m ≈ Rp 4 to 8 million /ha Assumption
Cost per paid hectare at 50 % of points paid Rp 4 to 8 m ÷ 0.5 = Rp 8 to 16 million /ha Illustration
Breakeven paid share at the low fee Rp 4 to 8 m ÷ Rp 15 m ≈ 27 % to 53 % Illustration
Points reaching selection 485 points ÷ 3 years (2027 to 2029) ≈ 160 a year Assumption on timing
Capturable revenue 2 points × Rp 3.3 billion = Rp 6.6 billion; 5 points × Rp 8.8 billion = Rp 44 billion a year Illustration
Java private-land need 107,086 ha − 28,000 ha = 79,086 ha; 107,086 ha − 8,500 ha = 98,586 ha Sourced inputs [6][10]

Weakest input: the fee, since no record of what developers pay for land work was found in repeated targeted searches, in Indonesian and English. Dollar figures use an assumed Rp 16,500 per US$.

The parties the business needs

  • Developers gain tenure before their deadlines (speed, quality) against an unknown fee: net unknown [9][2].
  • Small owners gain programme rent without selling or paying document costs (access, price) but give up farm use for decades at an unpublished rent: net unknown [10][13].
  • PLN gains more verified, contiguous hectares (quality, access) but has framed the register as cutting out intermediaries: net unknown [10].
  • Land offices and village heads save work on consolidated files (speed); village heads may lose informal influence over parcel deals: net unknown, not documented.

Routes considered

Folded into Ready-to-Lease Village Land Blocks: preparing several points in parallel to share organisers and equipment; one drone map with neighbour sign-offs instead of separate surveys; village self-screening of disputed plots; checks scaled to each parcel’s risk; a desk screen of spatial plans and tax maps; a few standard lease plans; and adding the line strip to the substation. A master lease through the village, or a cooperative form, are legal shapes the owners’ vehicle could take.

  • Faster bulk land certificates: depends on the land office working faster than usual, which a new company cannot control.
  • Bidders paying for verification before the tender, or a shared option fund among bidders: no sign developers would fund pre-tender work, and it raises tender-fairness questions.
  • Services for large landholders (paperwork, readiness certificates, tracks for mining or plantation land): those holders already control whole blocks and can register free or hire a consultant.
  • Rent tied to the tariff or to generation: adds verification cost and a dependency PLN may refuse, with no shown need.

Open questions

  • Will PLN accept a pooled owners’ vehicle organised by a paid service, and when will the next registration round open? PLN’s DPT Tanah team could answer.
  • What do developers pay for land work now? Former developer land teams could say.
  • What do PLN and the land office require per parcel, and what does that cost? Licensed surveyors and notaries could price one point.
  • At what rent would owners sign? A standard offer at one village would show.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *