When you screen land in mountain areas, on remote islands or in hot-spring towns, some projects die before the feasibility discussion even starts: the site has no public water main. On land outside a water-supply service area, however cheap the land and however fine the view, you cannot fix the room count or the construction schedule until you decide where the water will come from. And the choice of water source is also the entry point to a separate body of regulation, the Waterworks Act. This article looks at the water-source decision for sites without a mains connection on three levels: regulation, geography and cost.

Regulation — Once the Water Source Is Your Own, the Property Starts to Look Like a Water Utility

The moment you decide to drill a well or draw spring or stream water on a site without mains water, the property becomes subject to the Waterworks Act. Discovering this late in the business plan leads straight to a redone confirmation application and a delayed schedule. It is worth getting the categories right first.

Article 3(6) of the Waterworks Act defines a private waterworks as a waterworks other than one used for a water utility business, such as a private system at a dormitory, company housing or sanatorium, that either (1) supplies water needed for daily living to more than 100 people, or (2) has facilities whose maximum daily supply exceeds the threshold set by Cabinet Order (20 cubic meters). For accommodation, hotel guests are normally not treated as “residents”, so in practice the test that matters is (2), the 20-cubic-meter maximum daily supply.

If, on the other hand, the property connects to the mains and simply installs a receiving tank, it is a simplified private waterworks. That status depends on whether the tanks’ combined effective capacity exceeds 10 cubic meters, and the obligations are far lighter than for a private waterworks. The same “hotel water system” moves up a regulatory tier depending on whether the water source is your own.

In practice, the burden comes down to three things. The first is the pre-construction confirmation application. Article 32 of the Waterworks Act requires anyone who plans to build a private waterworks to obtain, in advance, the prefectural governor’s confirmation that the design meets the facility standards in Article 5. Preparing the design documents and waiting for the review have to be built into the schedule, and this runs on a separate track from building confirmation. It is easy to overlook that the same procedure applies when expanding or modifying an existing facility.

The second is ongoing water quality testing. The three-tier regime — color, turbidity and residual chlorine at least daily, the 9 basic items at least monthly, and all items every 3 months — effectively asks a hotel to operate like a small water utility. On top of that, the water quality standards amendment that took effect in April 2026 added PFOS and PFOA as standard items (combined value 50 ng/L), and private waterworks and small-scale municipal systems are now, as a rule, required to test for them once every 3 months. Items that used to be measured voluntarily against provisional target values are now mandatory. For anyone screening sites in 2026, this is a running cost that has only just been added.

The third is appointing a waterworks technical manager. A private waterworks must have one waterworks technical manager to handle the technical side of water management. At a small property, the choice is between having the general manager or facilities staff meet the qualification requirements and take on the role, or outsourcing it.

Geography — A “Market With Different Infrastructure Assumptions” of 2,120 Properties and 68,000 Rooms

With the rules in place, the next question is where such sites actually are and how many there are. Start with the national statistics. According to MLIT’s basic waterworks statistics, national water supply coverage at the end of FY2024 (as of March 31, 2025) was 98.3%, with 121.50 million people served. That breaks down into 119.63 million served by the main public supply, 1.52 million by small-scale municipal systems and 360,000 by private waterworks. Put the other way round, 2.11 million people out of a total population of 123.61 million are not served by any waterworks.

The differences between prefectures are significant. Coverage is lowest in Kumamoto at 89.8%, followed by Akita at 92.1%, Oita at 92.4%, Toyama at 93.9% and Ehime at 94.0%. Tokyo, Aichi, Osaka, Nara and Okinawa are at 100.0%. Kumamoto is known for its heavy reliance on groundwater, and sources other than the public supply are well established there for household use. Low coverage does not automatically mean an area is “behind”. Some areas show these figures because abundant high-quality groundwater makes laying water mains less necessary.

On that basis, we analyzed location conditions for a population of 18,105 properties and 1,221,815 rooms — operating accommodation properties tracked by MetroEngines Research with 10 or more rooms and known coordinates. 2,120 properties (11.7%) and 68,340 rooms (5.6%) fell into the mountain / low-density category. The group’s medians are 18 rooms, an elevation of 761 m and 147 residents within 1 km. The comparison group (the remaining 15,984 properties) has medians of 37 rooms, 17 m and 8,181 residents — clearly a completely different set of market conditions.

The distribution is extremely uneven. Nagano stands out with 500 properties and 15,693 rooms, 38.7% of its room stock. Its median elevation is 1,025 m, a direct reflection of how mountain resorts at around 1,000 m — Shiga Kogen, Hakuba, Yatsugatake, Kamikochi — form the core of the prefecture’s accommodation supply. Next come Niigata with 5,925 rooms, Gunma with 5,538, Fukushima with 5,080 and Hokkaido with 4,970. By share of prefectural room stock, the order is Nagano 38.7%, Gunma 29.1%, Niigata 22.1%, Yamanashi 21.9% and Fukushima 21.7% — all prefectures whose main tourism assets are hot springs and ski resorts.

Overlaying water supply coverage reveals two different patterns. Kumamoto (89.8% coverage), Oita (92.4%), Iwate (94.7%) and Gifu (95.0%) are prefectures where sources other than the public supply are common across the board, with large populations served by small-scale municipal systems and private waterworks. Nagano (99.0%), Gunma (99.6%) and Niigata (99.5%), by contrast, have high coverage but some of the largest populations served by small-scale municipal systems in the country: 82,001 people in Nagano, 78,826 in Niigata, 64,279 in Gunma and 68,301 in Yamanashi. In these prefectures, small-scale systems have historically covered mountain villages across wide areas, and many accommodation properties are probably connected to them. The situation is not “no water main” but “a water system other than the main public supply”, so land due diligence must always confirm whether the connection is to the main public supply or to a small-scale system. Small-scale systems operate at a small scale and are more exposed to consolidation and tariff revisions.

The estimated ADR column is worth a closer look. The median estimated contracted ADR for mountain / low-density locations is ¥14,900, well above ¥9,000 for the comparison group. The group is dominated by ryokan and resort hotels, where high-priced packages with meals are the main product. The less favorable the infrastructure assumptions, the higher the rate — the relationship is visible here. This is not a coincidence: being deep enough in the mountains that the mains do not reach is itself the source of scarcity value — scenery, quiet, hot springs — and that matters for the investment decision discussed later.

Slicing by elevation makes the relationship even clearer. In the 0–100 m band, only 8% of properties are in low-density locations, but the share rises to 33% at 100–300 m, 47% at 300–500 m, 49% at 500–800 m, 68% at 800–1,200 m and 96% at 1,200 m and above. Of the 325 properties and 11,084 rooms above 1,200 m, 311 properties and 10,559 rooms have fewer than 500 residents within 1 km. In other words, virtually all high-altitude accommodation supply sits away from settlements. Japan’s mountain resorts, one could say, are an industry built on conditions where public water mains are hard to extend.

On the map, the distribution stretches in bands along the Chubu highlands, the mountains of northern Kanto, the Ou Mountains running down the spine of Tohoku, and inland Hokkaido. The points near the coast are hot-spring towns on high ground or basins between mountains; almost none appear on the plains. The cost of laying water mains scales with the inverse of population density — an obvious structure that shows up directly on the map of accommodation supply.

Scale — How Many Rooms Make a “Private Waterworks”?

Next, we work backward from water volume to find the size at which a property with its own water source becomes subject to the private waterworks rules. The threshold is a maximum daily supply of 20 cubic meters.

There are two ways to set the unit water consumption. One is the maximum daily use used to design water supply equipment: public waterworks design standards set it at 2.0 cubic meters per day per hotel room and 1.0 cubic meter per day per business hotel room. On this design unit, a full-service hotel exceeds 20 cubic meters at 11 rooms and a business hotel at 21 rooms.

The other is building up from actual use. The Japan Association of Tourism Facilities reported, from a survey of about 100 properties nationwide, water use of 630 liters per guest (the average for a sample that includes many ryokan with large communal baths). At 2 guests per room and full occupancy, that is 1.26 cubic meters per room per day, which exceeds 20 cubic meters at 16 rooms. On an average day at an assumed 70% occupancy, it works out to 0.88 cubic meters per room per day, equivalent to 23 rooms. Because the test is based on maximum daily supply, it is applied on peak days, so in practice 16 rooms is a reasonable rule of thumb. Room count acting directly as a regulatory threshold is not unique to water; under the Fire Service Act, for example, sprinklers become mandatory at a total floor area of 6,000 m², which works out to roughly 103–240 hotel rooms.

Capex Estimate — A 150 m Deep Well Costs ¥11.35M for 20 Rooms and ¥20.75M for 100 Rooms

Now to the costs. Given the nature of the estimate, please read it with all assumptions disclosed. Well drilling and pumps can use published price ranges, but treatment and disinfection equipment, receiving tanks and the preparation of design and confirmation application documents vary widely with scale, water quality and terrain, and have no published market prices. Those items are values set by this article; real projects always need individual quotes.

On these assumptions, annual water use is 6,439 cubic meters for 20 rooms (25.2 cubic meters on a full day), 16,096 cubic meters for 50 rooms (63.0 on a full day) and 32,193 cubic meters for 100 rooms (126.0 on a full day). All of them exceed the 20-cubic-meter maximum daily supply.

Initial cost per room is ¥568k–¥680k at 20 rooms and ¥208k–¥230k at 100 rooms, a gap of nearly 3x, because drilling one well is a fixed cost largely independent of size (Figure 3). Herein lies the structural difficulty. Of the 2,120 mountain / low-density properties, 1,144 (54%) have 10–19 rooms, so supply is concentrated in the band where economies of scale help least. Moreover, estimated ADR by size band is almost flat — ¥15,200 for 10–19 rooms and ¥16,000 for 100+ rooms — so smaller properties do not command higher rates. The per-room burden of the water source weighs most heavily on small properties.

Over 20 years, the ranking reverses. For 100 rooms with a 1 km connection, the private source costs ¥62.51M–¥68.73M, while the mains connection costs ¥272.54M–¥475.82M. For 50 rooms it is ¥46.31M–¥51.56M for the private source against ¥143.77M–¥250.47M for the mains; for 20 rooms, ¥36.32M–¥40.9M against ¥66.51M–¥115.26M.

The reason is simple: a private water source has no volume charges. Mains water is billed by use, so costs rise in proportion as occupancy and guest numbers grow. A private source replaces that with an upfront investment and largely fixed annual costs for testing, the technical manager and maintenance. Only pumping electricity varies with volume, and even at 100 rooms it is ¥689k a year, about a third of the total.

Per occupied room-night, the picture is easier to grasp. Dividing by 20 years of occupied room-nights at an assumed 70% occupancy, the private source costs ¥122–¥135 at 100 rooms, ¥181–¥202 at 50 rooms and ¥355–¥400 at 20 rooms. A mains connection (1 km) costs ¥533–¥931 at 100 rooms, ¥563–¥980 at 50 rooms and ¥651–¥1,128 at 20 rooms. Against the median estimated ADR of ¥14,900 for mountain / low-density locations, the private source accounts for 0.8–0.9% at 100 rooms and only 2.4–2.7% even at 20 rooms.

The sign does not change in any of the three scenarios. Even in the combination least favorable to the private source (private source at the high end, mains at the low end), a gap of ¥25.61M remains at 20 rooms and ¥203.81M at 100 rooms. Over a 20-year horizon, the conclusion of this comparison holds up against the range of assumptions. That said, it assumes a 1 km connection; for a site next to a water main, the upfront cost on the mains side falls sharply, so the robustness of the conclusion does not carry over as is.

Even at a rate as low as ¥200 per m³, the private source is cheaper in all 15 cells. The reversal comes not from the rate but from a lower upfront cost on the mains side — that is, from a site next to a water main. Under the Table 3 base assumptions, once the connection distance falls below about 1.1 km for 100 rooms, about 0.8 km for 50 rooms and about 0.6 km for 20 rooms, the upfront cost of a mains connection drops below that of a private source, and from there the length of the holding period decides. In site screening, check the distance to the main before worrying about rate differences.

Risk — The Savings Come From Carrying the Risk Yourself

Over 20 years, a private water source is cheaper. But the difference is the price of taking on risks that a water utility would otherwise carry. Reading it as a pure cost difference leads to the wrong decision.

Quantity risk (drought). According to MLIT, droughts requiring intake restrictions occurred on 21 rivers nationwide in 2023. Droughts are spreading across wider areas regardless of season, and the Japan Water Agency notes that “water supply cuts and suspensions still occur nationwide in recent years, and in parts of Shikoku they have occurred in 8 or more of the last 30 years.” Changing rainfall patterns, with about 9.3 fewer rainy days per year than 100 years ago, are also part of the background. With a private source, a falling water table directly reduces the volume you can draw, feeding through to revenue in the form of restrictions on the communal bath or fewer rooms sold. With mains water, some of an intake restriction is absorbed by the utility’s adjustments; with a private source, there is no buffer.

Quality risk. Private waterworks must check residual chlorine daily and test monthly and every 3 months — which also means that if a test exceeds the standards, use must stop immediately and an alternative supply must be found. One case cited in Ministry of Health, Labour and Welfare materials involved spring water used at a hotel in which enterohemorrhagic E. coli O121 was detected; 37 people were infected and 16 developed mild food poisoning symptoms. Chlorine was normally added, but not on the day of the incident. Chlorine disinfection is not finished once the equipment is installed; the risk materializes the moment daily operation lapses. In addition, from April 2026 PFOS and PFOA were added to the water quality standards (combined 50 ng/L), with testing required every 3 months. PFAS levels in groundwater are hard to predict in advance, and adding treatment equipment after drilling reveals an exceedance is a risk that cannot be priced at the time of land acquisition.

And the flip side — self-sufficiency in a disaster. For all these risks, a private water source has a strength the mains lack. Unlike water pipes buried parallel to the ground, a well drilled vertically is less likely to break in an earthquake. One survey found that 95% of deep wells 30–300 m deep remained usable after the Great East Japan Earthquake. In the 2024 Noto Peninsula Earthquake, damage to water facilities caused prolonged outages, and in some affected areas well and spring water was voluntarily opened up for household use. In response, the Cabinet Secretariat’s Headquarters for Water Cycle Policy and MLIT issued “Guidelines for Groundwater Use in Disasters” in March 2025, recommending registration systems for disaster wells. The ability to keep operating during a water outage, or to serve as a local water point, is a clear asset from a BCP perspective. Known examples include Hotel New Otani (ホテルニューオータニ), which made its disaster-cooperation well drinkable, and Higo Bank (肥後銀行), which installed wells at 10 of its branches in Kumamoto that anyone can use in a disaster under an agreement. Hazard conditions and disaster-prevention Capex can also be quantified as a separate axis of site evaluation.

Investment Decision — Is Your Own Water Source a Cost, or Irreplaceable Backing?

Let us split the analysis so far by holding period. For the same asset, the meaning of a private water source flips depending on how long you expect to hold it.

In practical site screening, sticking to the right order is the most efficient approach. (1) Can you drill? (check whether the site is in a designated groundwater pumping restriction area or covered by an ordinance) → (2) Can you connect? (check whether the site is inside the service area, whether the supply is a small-scale system or the main public supply, the distance to the main and the municipality’s cost-allocation rules) → (3) How many rooms? (above 16 rooms, plan the schedule and organization on the assumption of a private waterworks) → (4) How much? (quotes for drilling and treatment equipment, and a 20-year comparison with the connection case). Working through the questions in this order avoids pushing ahead on land price and scenery alone, only to see the schedule collapse at the confirmation application stage. Before step (2), note that development permits and farmland conversion can add significant lead time, and that forest land has its own development permit requirements.

Finally, a word on the phrase “disadvantaged areas”. A site without mains water is not an “area left behind in development” but a market with different infrastructure assumptions. 68,340 rooms, 5.6% of the country’s accommodation supply, operate under those conditions, and many of them sell at higher rates than properties on the plains. Owning your water source is a choice to take on regulation, testing and equipment renewal in exchange for turning variable costs into fixed costs and gaining the self-sufficiency to keep operating during a water outage. Whether you deduct it as a cost or value it as backing for an irreplaceable location depends on your holding period and operating policy.

Summary

Sites without mains water are not an exception in Japan’s accommodation market. Behind the 98.3% water supply coverage figure, 2.11 million people use water from sources other than the main public supply, and on the supply side, 2,120 properties and 68,340 rooms are in mountain / low-density locations. In Nagano, 38.7% of the prefecture’s rooms are under these conditions.

With your own water source, a property becomes a private waterworks from 16 rooms (11 rooms on the design unit) and must file a confirmation application before construction, test water quality daily, monthly and every 3 months, and appoint a waterworks technical manager. In April 2026, PFOS and PFOA were added to the water quality standards, increasing the testing obligations. Assuming a 150 m deep well, the initial cost is ¥11.35M–¥13.6M for 20 rooms and ¥20.75M–¥23.0M for 100 rooms. Per room the gap is nearly 3x, and 54% of the qualifying properties are concentrated in the 10–19 room band, where economies of scale help least.

Even so, over 20 years a private source is cheaper, because volume charges drop to zero. For 100 rooms with a 1 km connection, the private source costs ¥62.51M–¥68.73M against ¥272.54M–¥475.82M for a mains connection. Those savings are the price of taking on drought and water-quality risk yourself — and also the price of the self-sufficiency to keep operating during a water outage. In site screening, work through the questions in order: can you drill, can you connect, how many rooms, and how much.

Related Articles

• Hotel Water & Sewer Costs Vary 2.1x by Municipality: ¥7.85mn NOI Gap
• Who Owns the Hot Spring? Water Rights and Ryokan Valuation