Dialysis is one of the few healthcare businesses in India with genuinely predictable demand. Patients on maintenance haemodialysis attend three times a week, every week, for years. Once a patient settles into a centre, they rarely move.
That predictability is why dialysis attracts entrepreneurs and nephrologists who have never built a healthcare facility before. It is also why the failure mode is so specific: centres rarely fail because demand did not appear. They fail because the water system was cheap, the third shift never happened, or the payer mix could not carry the fixed cost.
This article is a planning model, written for someone doing the arithmetic for the first time.
Every rupee figure here is an indicative planning range for current Indian market conditions, not a quotation. Verify equipment pricing with vendors, tariffs with the current notifications in your state, and rent and salaries with your own local market.
The short answer
A 10-station dialysis centre in India plans at ₹1.0 crore to ₹1.85 crore as a greenfield turnkey fit-out with new machines — roughly ₹10 to ₹18.5 lakh per station. On the planning model set out later in this article, a centre reaching 70 percent utilisation across three shifts returns its capital in approximately five years. At 90 percent utilisation, closer to two and a half. Below about 52 percent, it does not return it at all.
| Centre size | Built-up area | Indicative total (new machines) | Per station |
|---|---|---|---|
| 5 stations | 900 – 1,200 sq ft | ₹58 lakh – ₹1.05 crore | ₹11.5 – 21 lakh |
| 10 stations | 1,800 – 2,200 sq ft | ₹1.0 – 1.85 crore | ₹10 – 18.5 lakh |
| 20 stations | 3,200 – 4,000 sq ft | ₹1.85 – 3.3 crore | ₹9 – 16.5 lakh |
Smaller centres cost more per station because the RO plant, technical room, HVAC plant, reception and management layer are largely fixed. Specifying certified refurbished machines instead of new brings a 10-station centre down to roughly ₹77 lakh to ₹1.4 crore.
Setup cost, line by line
| Head | 5 stations | 10 stations | 20 stations |
|---|---|---|---|
| Civil works, interiors, flooring, drainage | ₹10 – 17 L | ₹18 – 30 L | ₹30 – 50 L |
| RO water plant, tanks, distribution loop | ₹6 – 10 L | ₹8 – 15 L | ₹13 – 23 L |
| Dialysis machines (new) | ₹22.5 – 40 L | ₹45 – 80 L | ₹90 – 155 L |
| Dialysis chairs or beds | ₹2 – 4.5 L | ₹4 – 9 L | ₹8 – 17 L |
| Bed head units, gas, vacuum, nurse call | ₹2 – 3.5 L | ₹4 – 7 L | ₹7 – 13 L |
| HVAC and ventilation | ₹4 – 8 L | ₹6 – 12 L | ₹10 – 20 L |
| Electrical, UPS, DG backup | ₹4 – 7 L | ₹6 – 11 L | ₹10 – 18 L |
| IT, dialysis software, CCTV | ₹1.5 – 3 L | ₹2 – 4 L | ₹3 – 6 L |
| Furniture, crash cart, reprocessing | ₹3 – 5 L | ₹4 – 7 L | ₹6 – 11 L |
| Statutory, licensing, professional fees | ₹1 – 2 L | ₹1 – 2 L | ₹2 – 4 L |
| Opening consumable stock | ₹2 – 5 L | ₹4 – 8 L | ₹6 – 13 L |
| Total | ₹58 – 105 L | ₹102 – 185 L | ₹185 – 330 L |
Three things move a quotation more than everything else.
Shell condition. Fitting out an existing hospital floor with power, water and drainage already in place cuts 25 to 35 percent off the civil line. A bare shell does not.
Machine specification. New volumetric-UF machines run ₹4.5 to ₹8 lakh each; certified refurbished units run ₹2 to ₹3.5 lakh. Refurbished is a legitimate way to open — provided you insist on documented service history, current-generation ultrafiltration control, five-year spares availability, and an AMC from someone who answers the phone.
Water system redundancy. A single RO skid is cheaper than a duplex skid with standby pumps. One membrane failure then cancels every session that day. For a centre that intends to run three shifts, that is not a luxury line item.
Full detail on layout, isolation and commissioning is on our dialysis unit setup service page.
The RO water plant: the most expensive place to save money
Here is the thing a first-time operator needs to internalise before anything else. A dialysis centre is a water treatment plant with a clinical floor attached.
A patient on thrice-weekly haemodialysis is exposed to roughly 20,000 litres of your treated water a year across a membrane a few micrometres thick, with no intestinal barrier in the way. Dialysis water is therefore engineered to limits an order of magnitude tighter than drinking water — around twenty-two chemical contaminants with specified maximum levels under the ISO 23500 series, a microbial ceiling of 100 CFU/mL with a 50 CFU/mL action level, and endotoxin at 0.25 EU/mL with a 0.125 EU/mL action level.
The water system is ₹8 to ₹15 lakh of a ₹1.25 crore project. Six to twelve percent. And it is where the compounding decisions sit.
| Specification choice | Capex delta | What it buys |
|---|---|---|
| Duplex softener instead of single | +₹0.5 – 1.0 L | Membranes stay protected during regeneration |
| Twin carbon beds in series with intermediate port | +₹0.4 – 0.8 L | Chloramine breakthrough detected before patients |
| Two-stage RO instead of single-pass | +₹1.2 – 2.5 L | Rejection margin and a fallback stage |
| Recirculating PP-R/PEX ring main instead of branched PVC | +₹1.5 – 3.0 L | No dead legs, no chronic biofilm |
| Standby pump set and duplex skid capability | +₹1.5 – 2.5 L | A membrane failure does not cancel the day |
| Sizing for target shift pattern, not opening occupancy | +₹1.0 – 2.0 L | The third shift stays available |
| UV steriliser plus endotoxin-retentive ultrafilter | +₹0.4 – 0.9 L | Final barrier before the loop returns |
| Hot-water sanitisation capability | +₹0.5 – 1.5 L | Disinfection that reaches everywhere water reaches |
Taken together, those eight decisions are most of the difference between the bottom and the top of the ₹8 to ₹15 lakh band.
Now the cost of getting them wrong.
Undersizing is the most expensive mistake, and the least visible. Each machine draws roughly 500 to 800 mL of treated water per minute during treatment — about 120 to 150 litres per four-hour session. Ten stations running three shifts need on the order of 500 to 750 litres per hour of permeate, and at 50 to 70 percent recovery that means 1,000 to 1,500 LPH of feed capacity, plus disinfection and rinse volumes. If you size the plant for the two shifts you open with, the third shift is not available later without replacing the skid. That third shift is worth roughly 180 sessions a month at 70 percent utilisation. On the contribution margin used later in this article, that is about ₹2.6 lakh a month — over ₹31 lakh a year, forgone permanently, to save ₹3 to ₹4 lakh once.
Biofilm remediation costs far more than doing the loop properly. A branched PVC tree with dead legs will eventually fail endotoxin testing and will not recover from disinfection, because biofilm in a stagnant branch is not reachable. The fix is repiping the loop in an operating centre: ₹4 to ₹9 lakh of work, one to three weeks of cancelled sessions, and — the part that actually hurts — patients transferred to a competitor. Dialysis patients are lifelong and sticky in both directions. They are hard to win and, once settled elsewhere, they do not come back. A two-week shutdown is not a two-week revenue loss; for some proportion of your patients it is permanent.
Chlorine and chloramine failures are a patient safety event, not a maintenance issue. Chlorine destroys RO membranes; chloramine passes straight through RO and causes haemolysis. The defence is twin carbon beds in series with a sampling port between them, sized for at least ten minutes of empty bed contact time, and a chloramine test before every treatment day. That defence costs under ₹1 lakh of capex and a few minutes of a technician’s morning.
Hard water reaching the membranes, because a single softener leaves them unprotected while it regenerates, means membrane replacement at ₹40,000 to ₹1.2 lakh a set, repeatedly.
If you cut one thing from a dialysis budget, cut a chair. Do not cut the water system.
The operating model: stations, shifts and patients
Everything in dialysis economics reduces to one equation:
Sessions per month = stations × shifts per day × operating days × utilisation
A standard haemodialysis session is four hours, and turnaround — machine disinfection, bay cleaning, patient changeover — takes 30 to 45 minutes. That is a cycle of roughly four and three-quarter hours:
| Operating day | Shifts achievable |
|---|---|
| 8 hours | 1 comfortable, 2 tight |
| 12 – 14 hours | 3 |
| 16 – 18 hours | 4 |
Three shifts is the planning target because it is the point at which the business works. Capital, rent and most manpower are committed at one shift. Going from two shifts to three raises capacity by 50 percent while adding only a technician layer and consumables. It is the single largest lever in the model.
What achieves three shifts is a patient base, not a decision. Do the arithmetic in patients: 10 stations, 3 shifts, 6 days a week is 180 session slots per week, and a maintenance patient uses three of them. Full three-shift operation therefore needs about 60 regular patients; 70 percent utilisation needs about 42.
Building that base takes quarters, not weeks. It comes from nephrologist relationships, hospital discharge referrals, scheme empanelment and being the closest reliable centre to where patients live. Most centres open on two shifts and reach three within 9 to 18 months. Plan on the slower end.
Revenue: payer mix drives the entire business case
There is no single “price of a dialysis session” in India. There is a spread of roughly three to one between the top and bottom of the market, and where your centre sits in that spread determines whether the numbers work at all.
| Payer | Indicative realisation per session |
|---|---|
| Private self-pay, metro corporate hospital | ₹2,500 – 4,500 |
| Private self-pay, standalone centre, tier-1/2 | ₹1,400 – 2,500 |
| Insurance / TPA reimbursed | ₹1,600 – 3,000 |
| PM-JAY, CGHS and state scheme empanelled | ₹1,000 – 1,800 |
| PMNDP or state PPP per-session contract | ₹900 – 1,600 |
These are directional planning bands. Scheme and tender rates are notified, revised periodically and differ state by state — read the current notification or tender document for your state rather than any published range, including this one.
The consequence is worth stating plainly: a standalone centre paying commercial rent, on a scheme-only payer mix, is very difficult to make work at typical Indian cost structures. The sensitivity table below shows why. Scheme volume is valuable as base load that fills your third shift; as the whole business it usually requires either a PPP structure where the government carries premises and utilities, or the procurement scale of a multi-centre network.
Operating costs
Consumables, per session
| Item | Indicative cost |
|---|---|
| Dialyser (single use) | ₹300 – 600 |
| Blood tubing set | ₹80 – 150 |
| Dialysate concentrate (acid + bicarbonate) | ₹60 – 120 |
| AV fistula needles (pair) | ₹40 – 90 |
| Saline, heparin, syringes, dressings | ₹60 – 130 |
| Gloves, PPE, surface disinfectant | ₹40 – 110 |
| Total, single-use protocol | ₹580 – 1,200 |
A controlled dialyser reuse protocol, where clinical policy and regulation permit, brings the per-session figure toward the lower end of that band. Volume purchasing does the rest — a 20-station centre buys materially better than a 5-station one, which is a real and underappreciated argument for scale.
Fixed costs, per month, 10-station centre
| Head | Indicative range | Model value |
|---|---|---|
| Dialysis technicians (4 – 6) | ₹0.8 – 1.8 L | ₹0.90 L |
| Nurses (2 – 3) | ₹0.5 – 1.2 L | ₹0.55 L |
| Nephrologist — retainer or per-session | ₹0.5 – 2.5 L | ₹0.70 L |
| Biomedical / technical support | ₹0.15 – 0.6 L | ₹0.20 L |
| Centre manager, front office, billing | ₹0.4 – 1.1 L | ₹0.55 L |
| Housekeeping | ₹0.2 – 0.5 L | ₹0.22 L |
| Rent, 1,800 – 2,200 sq ft | ₹0.5 – 3.5 L | ₹0.75 L |
| Power | ₹0.6 – 1.6 L | ₹0.85 L |
| Raw water and effluent | ₹0.1 – 0.4 L | ₹0.13 L |
| RO media, membranes, disinfectants (amortised) | ₹0.15 – 0.4 L | ₹0.18 L |
| Machine and RO AMC (amortised) | ₹0.35 – 0.7 L | ₹0.40 L |
| Biomedical waste disposal | ₹0.1 – 0.3 L | ₹0.12 L |
| Water testing programme | ₹0.05 – 0.15 L | ₹0.07 L |
| Software, telecom, insurance, admin, marketing | ₹0.2 – 0.5 L | ₹0.28 L |
| Total | ₹4.6 – 15.3 L | ₹5.90 L |
Two lines on that table deserve more attention than they usually get.
Power is a big number and consistently underestimated. A dialysis machine draws 1.5 to 2.5 kW during treatment, mostly for heating. Add the RO high-pressure and loop circulation pumps running continuously, HVAC across 2,000 sq ft, lighting and UPS conversion losses: budget 10 to 18 kWh per session at commercial tariffs. An outage mid-session is a clinical event, not an inconvenience, which is why DG and UPS are capital lines and not optional ones.
Water is cheap until it is not. Each session consumes 120 to 150 litres of treated water; at 50 to 70 percent RO recovery that is 200 to 300 litres of raw water, so a 10-station centre at three shifts draws well over 100 kilolitres a month. Where summer supply is unreliable, tanker water is a real line item and a real operational risk. Plumb the RO reject to flushing or landscaping while the walls are still open.
A worked payback model
What follows is illustrative planning arithmetic, not a projection of returns and not financial advice. It is a structure for building your own model with your own numbers. No two catchments, tariff schedules or rent markets are alike, and nothing here should be read as a promise of outcome.
Assumptions
| Assumption | Value used | Basis |
|---|---|---|
| Stations | 10 | 9 general + 1 isolation |
| Capital cost | ₹1.25 crore | Midpoint of the ₹1.0 – 1.85 crore band |
| Machine specification | New | Refurbished would lower capex materially |
| Shell | Leased, 2,000 sq ft, tier-2 city | Rent in fixed costs |
| Shifts per day | 3 | 12 – 14 hour operating day |
| Operating days per month | 26 | Six-day week |
| Nominal monthly capacity | 780 sessions | 10 × 3 × 26 |
| Steady-state utilisation | 70% | Reached in month 15 – 18 on a realistic ramp |
| Sessions per month at steady state | 546 | 780 × 70% |
| Blended realisation per session | ₹2,200 | Mix weighted toward private and insurance |
| Consumable cost per session | ₹750 | Single-use protocol, volume purchasing |
| Contribution per session | ₹1,450 | ₹2,200 − ₹750 |
| Monthly fixed operating cost | ₹5.90 lakh | From the table above |
| Excluded from this model | Tax, depreciation, interest, working capital | Pre-tax EBITDA basis only |
Steady-state month at 70 percent utilisation
| Line | Amount |
|---|---|
| Sessions | 546 |
| Revenue | ₹12.01 lakh |
| Less: consumables | ₹4.10 lakh |
| Contribution | ₹7.91 lakh |
| Less: fixed operating cost | ₹5.90 lakh |
| Monthly EBITDA | ₹2.01 lakh |
| Annual EBITDA | ₹24.1 lakh |
| Simple payback on ₹1.25 crore | ≈ 5.2 years |
Sensitivity
This is the part worth spending time on, because dialysis economics are steeply non-linear. Fixed costs do not move with volume, so every session above breakeven contributes ₹1,450 straight to EBITDA.
| Scenario | Sessions/month | Monthly EBITDA | Annual EBITDA | Simple payback |
|---|---|---|---|---|
| 50% utilisation | 390 | −₹0.24 lakh | Negative | None |
| 70% utilisation | 546 | ₹2.01 lakh | ₹24.1 lakh | ≈ 5.2 years |
| 90% utilisation | 702 | ₹4.28 lakh | ₹51.4 lakh | ≈ 2.4 years |
| 70% but scheme-heavy mix (₹1,500 blended) | 546 | −₹1.81 lakh | Negative | None |
| 70% at two shifts only (520 nominal) | 364 | −₹0.62 lakh | Negative | None |
Those five rows contain the whole business.
Breakeven sits at roughly 407 sessions a month — about 52 percent of three-shift capacity. Below that the centre loses money regardless of how well it is run clinically.
Going from 70 to 90 percent utilisation more than doubles EBITDA and halves payback. The last 20 percent of capacity is worth more than the first 50, because the fixed cost is already paid.
A two-shift centre at good utilisation still loses money at these assumptions. This is why RO plant sizing matters so much: if the water system cannot support a third shift, the model never reaches the row where it works.
Payer mix can invalidate the case on its own. The same centre, the same utilisation, the same staff, is profitable at ₹2,200 blended and loss-making at ₹1,500. Nothing operational changes. Only the mix.
Build this table for your own centre before you sign a lease.
Risks and failure modes
The ramp is slower than the plan. Patients are lifelong and loyal, which cuts both ways — they are difficult to win from an incumbent. At 30 to 40 percent utilisation, this model burns ₹1.4 to ₹2.5 lakh a month. Budget 12 to 18 months of operating loss cover — ₹20 to ₹30 lakh — on top of capex. Running out of cash in month nine of an eighteen-month ramp is the most common way a viable centre dies.
Scheme reimbursement is slow. Empanelled and PPP business commonly settles on cycles measured in months, while consumable suppliers work on 30-day terms. At a 55 percent scheme mix on ₹12 lakh of monthly revenue, a three-month receivable cycle locks up around ₹20 lakh of working capital — a separate number from your loss cover, and one that grows with you.
Technician attrition. Dialysis technicians are scarce and mobile. In a four-technician roster, one resignation removes a shift’s coverage. Over-hire by one, cross-train nurses, pay above the local market — all cheaper than cancelled sessions.
Water quality failure. Covered above, and worth repeating in business terms: it is the only operational failure that can close a dialysis centre outright.
Machine downtime. One machine down for a week at three shifts costs about 18 sessions and ₹26,000 of contribution — small per event, significant when chronic. Carry one spare machine per eight to ten stations, and contract AMC on machines and the water system together.
Nephrologist dependence. In many centres, referrals follow one consultant. If the relationship is informal and undocumented, so is your patient base.
PMNDP and PPP: what participation actually means
The Pradhan Mantri National Dialysis Programme, run under the National Health Mission, provides free haemodialysis to eligible patients at government facilities through public-private partnership. It has been extended along a hub-and-spoke pattern, with district hospitals functioning as hubs supporting dialysis at sub-district and community health centre level, plus a peritoneal dialysis component for patients who cannot travel three times a week.
The general structure is consistent even where commercial terms differ by state. The government provides premises, power, water and patient flow. The private partner provides machines, the RO plant, consumables and manpower, and is reimbursed per session. States tender the operating contracts, so station counts, isolation provision, water testing frequency, uptime requirements and penalties are defined in the tender document rather than negotiated.
For an operator, this changes the economics in three directions at once.
It removes the hardest costs. Civil capex largely disappears. Rent, power and water leave the fixed cost table — in this model, roughly ₹1.7 lakh a month of the ₹5.90 lakh. Patient acquisition, the thing that kills private centres in year one, is not your problem.
It fixes the ceiling. Per-session realisation is set by tender and does not respond to service quality. Your only levers are volume, procurement cost and manpower efficiency.
It transfers risk in a specific direction. Uptime and water compliance become contractual obligations with financial consequences, so your backup power and water redundancy specification is driven by the contract rather than by preference — one more reason the RO plant is the wrong place to economise.
The pattern that follows is that PPP dialysis tends to work at network scale rather than as a single centre: procurement leverage across many sites, a shared biomedical and water engineering team, and pooled technician cover are what turn a fixed tariff into a workable margin. A first centre is usually better served by a mixed payer model, using scheme volume to fill capacity that private demand alone would leave idle.
Hub-and-spoke matters for planning too. If your catchment sits within a hub’s referral shed, a satellite centre may be the more capital-efficient entry — smaller station count, shorter travel for patients, and the hub handling complications.
Getting the sequence right
The decisions that set a dialysis centre’s economics for a decade are almost all made in the first three weeks, before anything is built.
- Model your catchment in patients, not sessions. Three shifts needs roughly six regular patients per station. Ask whether they exist.
- Decide your target payer mix before signing a lease — it determines what rent you can afford.
- Size the water system for the shift pattern you intend to reach, not the one you open with.
- Budget operating loss cover and scheme working capital as separate lines from capex.
- Build the sensitivity table above with your own numbers, and look hardest at the rows where it fails.
RayMedico Projects delivers turnkey dialysis unit setups from Pune — source water analysis and station-count modelling, layout and engineering, civil and MEP, the RO plant and distribution loop, equipment, and water validation with a NABH-ready documentation pack. If the unit sits inside a larger facility, our companion article on hospital construction cost in India covers the wider project economics.
For an assessment against your own catchment, shell and target shift pattern, request a quote.
About this article
Written by RayMedico Projects Editorial Team for RayMedico Projects. Cost figures, timelines, and specifications given here are indicative planning ranges drawn from current Indian market conditions — they are not quotations. Accreditation and statutory requirements are revised periodically; verify current requirements with the relevant authority for your project. For a project-specific assessment, request a quote.