Could a local company keep Semarang’s flood pumps ready for the rain?

A drainage pump in a coastal city does one job: it lifts rainwater and seawater out of low streets and pushes it into a river or the sea. In Semarang, the capital of Central Java, the city government owns these pumps and pays for them out of its public budget. In January 2026 its public works department, DPU, said it had 220 ready for the rainy season: 119 fixed pumps in pump houses and 101 portable ones [16]. On 19 November 2026 the head of the new authority for Java’s northern coast said the ground there has already sunk by 20 cm, and the news site that reported him headlined that Semarang’s airport is slowly going under [1]. In a city like that, a pump that does not start is a flooded street.

Pumps do fail to start. When inspectors from Ombudsman RI, the national public-services watchdog, visited the Kali Tenggang pump house in February 2025, they found that only four of its six pumps were running: two were still under repair, and the station had also had trouble getting diesel delivered [11]. That is one station on one day, not the state of the city today. But it shows what keeping a pump ready really involves: a working machine, fuel in the tank and an intake that is not blocked by rubbish, all at the same station at the same time.

We asked whether a new local company could sell Semarang exactly that: a contract to keep each pump ready before the rain comes. The answer turns on something the city already buys.

The city already pays a pump maker to keep its pumps working

Semarang does not leave its pumps to chance. The city’s pump operating units (UPTD, the department’s local field offices) are formally responsible for operating, maintaining and monitoring the pumps in their area [17]. The 2024 accounts the city posted online record about Rp14.2 billion spent on running and maintaining pump stations that year, a figure for the whole programme that the accounts do not split into staff, fuel, parts and contractors [18].

The department also hires specialists. In 2026 it advertised a contract to maintain the pumps at Kali Semarang, one of the city’s main drainage rivers, and asked bidders for a letter of support from Grundfos, the pump manufacturer [5]. The city’s procurement site shows that PT. GRUNDFOS Pompa, the manufacturer’s Indonesian company, won it, at a contract price of about Rp7.8 billion (about US$470,000) [22] [23]. The short statement of work says the job is to keep the pumps and their equipment at that river working, over 180 days [24]. It does not say how many pumps are covered, so the figure cannot be turned into a price per pump. The city’s 2025 standard price list gives one more reference point: it lists the servicing and repair of one large Grundfos drainage pump model at Rp49 million per job [21].

Local firms already sell this kind of work too. PT Adhirajasa Bangun Mandiri, a Semarang contractor, advertises repairs and maintenance of flood pumps and lists public pump-station projects among its past work [6] [20]. PT Arsindo, a pump-service company based near Jakarta, advertises preventive-maintenance contracts and a 6 to 12 hour emergency response in Semarang [3] [19], and lists daily rates for pump technicians and monthly pump rentals [2]. These are prices the companies advertise, not what the city pays. Grundfos also has a factory in Semarang, but it makes borehole pumps for groundwater rather than serving the drainage stations [14].

So the money is real and buyers already spend it. What the evidence cannot yet show is why pumps fail. One explanation is that the work is split: city crews run the pumps, specialists repair them and diesel is ordered separately, so no one is answerable for a pump being ready. We could not confirm that. The public record shows that failures happened, but not their causes, and not whether the 2026 contract has already fixed them.

How one pump would be looked after under contract

The business we examined is the Semarang Pump Readiness Service: a local company that takes responsibility for an agreed set of the city’s fixed pumps, whatever their brand, and sells the department one annual fee for keeping each of them ready. One pump under the contract would be handled like this, as we propose it (no such contract exists):

One pump-year under the proposed contract

  1. 1. Pump named
    The department picks the pump and shares its records and recent job orders
  2. 2. Pre-rain check
    Company technicians test it, check the intake and the fuel order, and price any defect
  3. 3. Repairs
    The company fixes what it can and pays specialists for the rest from its allowance
  4. 4. Operation
    City operators run the pump; approved fuel orders go to a paid supplier
  5. 5. Payment
    The department checks the job records and pays the annual fee
How the proposed company would look after one city pump; no such contract exists yet.Based on sources [11] [17] [24] and our own reasoning.

The city keeps what it has today: its staff operate the pumps and its budget pays for the diesel and electricity they burn. The company adds paid technicians, a repair roster for the wet months and one record of every job, so that one coordinator at the company sees that a pump’s repair, its fuel order and its intake check are all done, instead of the city chasing three different people. It sells finished, checked work, not a promise to keep streets dry, because no maintenance contract can control how much rain falls or how well the canals drain.

Related Business Opportunity:  Could local soybean deliveries cut costs for Central Java’s tofu makers?

Why would the department switch? Only if the company does work the city’s crews and its existing suppliers leave undone (a repair still pending when the rain starts, a fuel order placed too late) at a price no higher than getting the same work done another way. Because nobody has published what the city pays per pump today, we cannot yet say whether our assumed price below is higher or lower than that. Finding out is the first test, and the 2026 Grundfos contract is the comparison the company must beat.

A crew needs about 22 pumps to pay for itself

The economics decide whether this is a business or a consultancy. On our own assumptions, a cluster of 30 comparable pumps at Rp60 million (about US$3,600) per pump per year would bring in Rp1.8 billion. Two kinds of cost come out of that. Some are shared by the whole contract, about Rp860 million a year: four permanent technicians, four more for the five wettest months, vehicles, tools, insurance, management and the cost of bidding. The rest grows with each pump: the model sets aside Rp20 million per pump-year from the fee for routine work and for paying outside specialists for repairs, about Rp600 million for 30 pumps. Major rebuilds would be billed to the city separately. Total cost is about Rp1.46 billion, leaving about Rp340 million before financing and tax. Only the technicians’ pay floor is sourced: Semarang’s 2026 minimum wage is about Rp3.7 million a month [15]. The price, the crew size and the workloads are assumptions. Each pump leaves Rp40 million after its own allowance, so the shared Rp860 million is covered at about 22 pumps.

The weakest input is repairs. The model sets aside enough for about seven of the Rp49 million servicing jobs from the city’s price list across 30 pumps in a year; about 14 such jobs would wipe out the margin [21]. How often Semarang’s pumps need that work is not public.

The first market is one buyer. If the department contracted out all 119 fixed pumps this way, at an assumed Rp40 million to Rp96 million per pump-year, it would spend roughly Rp4.8 billion to Rp11.4 billion a year. A newcomer that won 20 to 30 percent of them would have about 24 to 36 pumps and Rp1.4 billion to Rp2.2 billion in annual revenue at the central price, enough for one small team. But the department’s own online dashboard lists only 78 pump machines across 42 pump houses [7], and the city has not published which count is current. On the lower count, the same share gives 16 to 23 pumps, below the size that pays for the crew. Beyond Semarang, Jakarta has the same job on a larger scale, with 602 fixed pumps in June 2025 [13], but it would need a new contract and a new local team.

Semarang’s other big pump owners are not obvious customers. The airport was reported preparing 58 pumps, and its operator is building a 700-metre dike and clearing channels on the airfield [9] [10], and the port operator Pelindo describes its own storm preparations at Tanjung Emas [12]. Their flooding calls for civil works as much as pump care, and neither has published what it spends on maintenance.

What would stop it

The strongest objection is that the job may already be done. Grundfos holds the Kali Semarang contract, Adhirajasa can quote the same bundle locally, and Arsindo can offer faster response under contract. Any of them could add fuel and intake checks to what they already sell, and nothing protects a coordinator’s schedule from being copied. The city may also be unable to buy a cross-brand package from a new supplier without a manufacturer’s letter for each brand [5].

The cheapest next step costs a few requests, not a fleet: obtain the detailed work list, pump count and invoices behind the Rp7.8 billion contract; the department’s own cost of the same work; one priced annual offer from a local firm; and recent failure records from a few stations. If those show that the existing providers already keep the pumps ready at a lower cost, or that too few pumps are open to a new supplier to keep a crew busy, this version of the business should be dropped. Engineers and procurement officers at DPU, pump technicians and fuel suppliers in Semarang could answer these questions.

In short: documented failures, a verified pump-maintenance contract and a published repair price show that Semarang pays for working pumps. Our cost model suggests a contract of about 30 pumps could support a small local team. Unknown, and decisive: whether there is preventable work the current providers leave undone, whether the department can buy this service from a newcomer, and what the city pays per pump today.

What has to be true

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

  1. DPU will pay about Rp60m per pump-year for the specified work instead of buying the same work more cheaply internally or from its existing agents.
    Not yet tested · DPU Kota Semarang
    Verified contracted purchase Rp7.8267bn for 180-day maintenance scope; pump count, detailed tasks and actual invoices unknown. This supports existing demand for maintenance but not the proposed annual price. [5] [18] [23] [24]
    Would break if: Its current same-scope cost or a credible incumbent quote is lower without a measurable readiness disadvantage.
    Cheapest test: Obtain the detailed OEM scope, covered pump count and invoices behind the Rp7.827bn contract; compare the Rp60m annual offer with matched internal and local quotes.
  2. Procurement officials can and want to buy a cross-brand readiness package that the new supplier can qualify to deliver.
    Not yet tested · DPU procurement officials
    [5] [17]
    Would break if: No acceptable budget, package, qualification or payment route exists for this supplier and scope.
    Cheapest test: Have the PPK review one scoped offer and required qualifications; obtain written route/eligibility clarification before hiring.
  3. Coordinating repairs, fuel checks and intake work will prevent enough recurring outages to justify an additional contractor.
    Not yet tested · DPU Kota Semarang
    [11] [16] [17]
    Would break if: Recent work orders show the same failures already resolved or primarily caused by civil drainage constraints outside this service.
    Cheapest test: Compare the latest year of outage and work-order logs for several stations; identify preventable, still-unowned tasks and their current costs.
  4. Principals or authorised specialists will supply the required parts, permissions and repair work on terms that fit the service price.
    Not yet tested · OEM principals / authorised specialist suppliers
    [5] [14] [20] [23] [24]
    Would break if: Required authorisation is unavailable or specialist charges consume the proposed margin.
    Cheapest test: Request written supply/authorisation and fixed-scope specialist quotes for the actual pump register.
  5. Qualified crews will accept the budgeted pay and commit enough simultaneous flood response capacity.
    Not yet tested · Local permanent and seasonal technicians
    [15]
    Would break if: Credible recruitment and duty-roster quotes exceed the cost allowance or leave peak jobs uncovered.
    Cheapest test: Price the actual duty roster with local qualified technicians, including premiums, overtime and five-month seasonal commitments.
  6. Pump operators will accept the reporting interface without unpaid extra work or loss of operating control.
    Not yet tested · UPTD pump operators
    [17]
    Would break if: The handoff adds uncompensated duties, conflicts with authority or delays existing operations.
    Cheapest test: Review one work-order and reporting handoff with UPTD staff; time it and allocate every additional task to a paid contractor role.
  7. A fuel partner can deliver paid orders to assigned stations before and during the relevant flood periods.
    Not yet tested · Fuel delivery partner
    [11]
    Would break if: The delivery route, lead time, storage/access approval or emergency price prevents fuel readiness within the budget.
    Cheapest test: Obtain a supplier’s site-specific delivery and contingency quote with agreed ordering responsibility; no live fuel trial needed first.
  8. At least roughly 22 similar pumps can share the proposed service team, with a 30-pump contract leaving enough allowance for repairs.
    Not yet tested · DPU Kota Semarang and service contractor
    2025 municipal standard lists a specific Grundfos service/repair at Rp49m per unit, not per year; recurrence and scope overlap are unknown. [15] [16] [21]
    Would break if: A site-by-site work plan needs more staff or repairs than the price supports, or fewer than 22 paid units can be assembled.
    Cheapest test: Build a priced route, inspection and response roster from the actual asset list, and compare covered work with the current provider scope.
Related Business Opportunity:  Could an outside team handle Indonesia's child-safety filings for mid-size apps and online shops?

Research Notes

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

All amounts are in Indonesian rupiah. Dollar equivalents use an assumed Rp16,500 per US dollar. Numbers 4 and 8 are not used in this list; sources are numbered as in our research record.

Who already provides this work

Provider What it offers Price evidence In Semarang’s city stations?
PT. GRUNDFOS Pompa Pump maker’s own service Contracted to keep Kali Semarang’s pumps working for 180 days [23] [24] Rp7.83 billion contract Pump count not published [22] [23] Yes [23]
PT Adhirajasa Bangun Mandiri Local flood-pump service Rental, operation, repair and public station projects [6] [20] No fee published A same-scope quote is needed Yes [20]
PT Arsindo Multi-brand maintenance contracts Preventive contracts, emergency repairs, rentals [3] [19] Advertised day rates Technician about Rp1.5 million a day; rentals Rp4.8–27.8 million a month [2] Unknown: advertises Semarang coverage, no city contract found [19]
Tehnik Optima Pompa Grundfos factory partner Makes borehole pumps in Semarang [14] No service tariff Unknown [14]
City pump crews (UPTD) In-house operation Operate, maintain and monitor pumps [17] Rp14.2 billion programme 2024 operation and maintenance, all costs [18] Yes [17]
River-basin agency (BBWS) and airport, port owners Their own stations Separate owners and budgets [9] [12] No maintenance spend published Not the city’s buyer

Economics workings (our illustration)

Item Calculation Status
Revenue, 30 pumps 30 × Rp60 million = Rp1.8 billion a year Assumed price
Four permanent technicians Rp266.5 million a year (minimum wage [15] × 1.5 loading, 13 months) Wage sourced; loading assumed
Four seasonal technicians Rp111.1 million (five months) Assumed
Tools and vehicles Rp120 million Assumed
Administration, supervision, insurance Rp180 million Assumed
Overtime and contingency Rp120 million Assumed
Winning the first contract Rp60 million Assumed
Work and repair allowance 30 × Rp20 million = Rp600 million Assumed
Total cost Rp857.6 million + Rp600 million ≈ Rp1.46 billion Calculated
Surplus before finance and tax Rp1.8 billion − Rp1.46 billion ≈ Rp342 million (19%) Calculated
Break-even size Rp857.6 million ÷ (Rp60 million − Rp20 million) ≈ 22 pumps Calculated
Repair sensitivity Rp12 million specialist share × 30 = Rp360 million ÷ Rp49 million ≈ 7 jobs [21]; about 14 such jobs erase the surplus Rate sourced; frequency unknown
Working capital 90 days of cost ≈ Rp364 million before the first payment Assumed payment delay

Doubling technicians’ pay to three times the minimum wage removes the surplus. A pump-for-pump comparison with the Grundfos contract needs its pump count and work list, which the published documents do not give [24].

Related Business Opportunity:  Could a village builder rebuild flood-hit homes in Aceh for the Rp70 million each family gets?

First market

Step Calculation Status
Buyers 1 city public works department Sourced [17]
Fixed pumps 119 (dashboard: 78) Sourced [16] [7]
Price per pump-year Rp40 million to Rp96 million, central Rp60 million Assumed
All fixed pumps 1 × 119 × Rp40–96 million = Rp4.76–11.42 billion a year (78 pumps: Rp3.12–7.49 billion) Calculated
A newcomer’s share 20–30% of 119 ≈ 24–36 pumps × Rp60 million = Rp1.44–2.16 billion a year (78 pumps: 16–23, Rp0.96–1.38 billion) Assumed share
Later market Jakarta: 602 stationary pumps at 205 locations [13] Sourced count; access untested

Routes considered

Folded into the Semarang Pump Readiness Service: fuel and intake checks at each visit, planned maintenance outside peak rain, a core crew with seasonal technicians, combined inspection visits, weather-triggered fuel orders, one work-order record across mechanics, fuel and intake, and keeping city operators in charge of running the pumps.

  • Shared standby pump pool. The same spare pumps cannot answer several floods at once, and the city already owns 101 portable pumps [16].
  • Fast multi-brand repair team. Arsindo already advertises 6 to 12 hour response in Semarang [19] and Adhirajasa is local, so speed alone is not a new advantage.
  • Spare-parts hub. No source shows that waiting for parts is what keeps pumps down, and holding stock is costly.
  • City-wide pumping-capacity guarantee. Pumping power in one district does not drain another, so a city-wide total cannot be promised.
  • Service for the airport and port. Both run their own preparations and civil works [10] [12], and no record shows a maintenance gap there.
  • On-call freelance technicians only. Without a committed crew, no one keeps station knowledge or can answer several floods at once.

Open questions

  • How many pumps does the Grundfos contract cover, what work does it list, and what has been paid? (DPU procurement officers.)
  • What does the department spend on the same work through its own crews? (DPU and UPTD engineers.)
  • What would a local firm charge for a year of this service? (Adhirajasa, Arsindo or other Central Java pump specialists.)
  • What failed at stations in the last two rainy seasons, and why? (UPTD operators.)
  • Can a new supplier qualify for a cross-brand contract without each manufacturer’s letter? (The department’s procurement officer.)

Sources

1. Tanah Pantura Turun 20 Cm, Bandara Semarang Perlahan Tenggelam (detikFinance, 19 Nov 2026) ↩

Opening context: subsidence of 20 cm along Java's north coast and the airport headline. An official's claim, not a diagnosis of pump maintenance. Access: full.

2. Sewa Pompa Industri – Harga Rental 2026 (PT Arsindo, pricelist 2026 R02) ↩

Arsindo's advertised 2026 rental and manpower rates (operator about Rp1.5 million a day, helper about Rp1 million a day; pump rental Rp4.8 million to Rp27.8 million a month before tax and transport). Seller prices, not paid prices. Access: full.

3. PT Arsindo — pump services, preventive maintenance and 24/7 repair contracts ↩

Arsindo's advertised preventive-maintenance contracts and repair service. Seller page. Access: full.

5. LPSE Kota Semarang — Pemeliharaan Pompa Kali Semarang (Paket Ulang), APBD 2026 ↩

The 2026 Kali Semarang pump-maintenance package: a buyer's estimate (HPS) of about Rp8 billion and the requirement for a Grundfos support letter. HPS is the buyer's estimate, not money paid. Access: full.

6. Jasa Sewa Pompa Banjir (PT Adhirajasa Bangun Mandiri, Semarang, 24 Feb 2023) ↩

Adhirajasa's flood-pump rental and operation service in Semarang (24 February 2023). Seller page. Access: full.

7. Smart Infrastruktur PU Semarang — Fasilitas Rumah Pompa (DPU Kota Semarang) ↩

DPU's online asset dashboard: 42 pump houses and 78 pump machines. Its survey and period fields (2021, 2026) make the date of the count unclear; read the counters and condition table only. Access: section.

9. Musim Hujan, Bandara Ahmad Yani Semarang Siapkan 58 Pompa Penyedot Banjir (IDN Times, 16 Dec 2024) ↩

Airport flood preparations, including 58 pumps (16 December 2024). Access: full.

10. Antisipasi Banjir, Bandara Jenderal Ahmad Yani Semarang Bangun Tanggul 700 Meter (Angkasa Pura Indonesia, 24 Feb 2025) ↩

The airport operator's 700-metre dike and airside channel clearing (24 February 2025). Access: full.

11. Ombudsman Jateng Pantau Rumah Pompa Kali Tenggang (Ombudsman RI, 7 Feb 2025) ↩

Ombudsman inspection of the Kali Tenggang pump house, 7 February 2025: four of six pumps working, two under repair, an earlier fuel-supply problem. Access: full.

12. Pelindo Siapkan Antisipasi Potensi Cuaca Buruk di Pelabuhan Tanjung Emas (Pelindo press release, 30 Dec 2022) ↩

Pelindo's storm preparations at Tanjung Emas port (30 December 2022). Access: full.

13. Update Pengerukan dan Jumlah Pompa Pengendali Banjir di Jakarta per Juni 2025 (Dinas SDA DKI) ↩

Jakarta's 602 stationary flood pumps at 205 locations as of June 2025. Section read: the inventory paragraph. Access: section.

14. Grundfos opens SP Sub Factory in Semarang with Tehnik Optima Pompa (Grundfos, 8 June 2023) ↩

Grundfos's submersible (SP) borehole pump factory in Semarang with Tehnik Optima Pompa (8 June 2023). Seller page. Access: full.

15. Central Java PPID: 2026 minimum wages, 24 December 2025 ↩

Semarang city's 2026 minimum wage, Rp3,701,709 a month. Access: full.

16. ANTARA: Semarang prepares 220 pumps, 5 January 2026 ↩

Semarang's 220 pumps (119 fixed, 101 portable) readied in January 2026. Access: full.

17. DPU Semarang PPID: duties of eastern flood-pump operating unit, 24 April 2026 ↩

The duties of a city flood-pump operating unit (UPTD): operation, maintenance and monitoring. Access: full.

18. Semarang published 2024 APBD accountability document, numbered 5/2025; cover marked RANCANGAN ↩

2024 city accounts: about Rp14.153 billion realised on pump-station operation and maintenance for the whole sub-programme. Only the cover and expenditure table were read; the posted document's cover is marked RANCANGAN (draft). Access: section.

19. PT Arsindo: emergency pump service, undated, accessed 19 November 2026 ↩

Arsindo's advertised emergency dispatch times: 6 to 12 hours for Semarang, and guaranteed response for maintenance-contract clients. Seller page. Access: full.

20. PT Adhirajasa Bangun Mandiri: pump services and historical project portfolio, undated, accessed 19 November 2026 ↩

Adhirajasa's pump services and project portfolio. Seller page, undated. Access: full.

21. Semarang Mayor Regulation 36/2025, amendment to 2025 standard unit prices: pump repair/service schedule ↩

Semarang's 2025 standard unit prices: servicing and repair of a Grundfos 450 KPL 306T3 drainage pump at Rp49 million per unit. A standard price per job, not an annual tariff or an invoice. Only the cover and that rate table were read. Access: section.

22. LPSE: Kali Semarang 2026 winner, offer and negotiation ↩

Procurement result: PT. GRUNDFOS Pompa's offer of Rp7,994,582,304 and the negotiated price. Access: full.

23. LPSE: Kali Semarang 2026 contracted winner and contract price ↩

Procurement result: contracted winner PT. GRUNDFOS Pompa, contract price Rp7,826,695,470. Actual payments were not found. Access: full.

24. DPU: Uraian Singkat Pemeliharaan Pompa Kali Semarang, APBD 2026 ↩

Short statement of work: functional maintenance of the pumps and equipment at Kali Semarang (UPTD Tengah II), 180 calendar days from the work order. No pump count or detailed quantities. Access: full.

Comments

Leave a Reply

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