Work out your mri machine power consumption in seconds: enter the wattage on your mri machine's label, its hours of use per day and your electricity rate, then click Calculate. You get your daily, monthly and yearly cost and total kWh consumption straight away. Those same results also cover how much electricity an mri machine uses. Also see how much energy does a desktop computer use.
If you have ever wondered how much electricity a hospital's imaging suite burns through, the answer starts with MRI machine power consumption: a single magnetic resonance imaging scanner typically draws well over 100,000 kWh of energy every year, and a large share of that is used while no patient is even on the table. This guide breaks the energy consumption down by operating mode, shows a worked calculation you can repeat with your own numbers, and explains where the savings are.
The honest answer is that it depends on the field strength, the gradient and RF amplifier hardware, the cooling design and, above all, how the MRI scanner is scheduled. A modern 1.5 T MRI unit typically peaks somewhere between 40 and 60 kW while it is acquiring images, but it never drops to zero. Because the superconducting magnet must stay cold, the machine keeps drawing power overnight, on weekends and during holidays.
That is why the useful figure is not a single wattage but a profile. Published on-site measurements of MRI systems consistently show three operating states, and the split between them explains most of the total energy usage:
Studies of hospital MRI machines report that stand-by and idle time together can account for well over half of the total energy consumption in MRI, which is the single most important thing to understand before trying to cut the bill.
Seeing the three states side by side makes MRI energy consumption easier to reason about. The table below uses a hypothetical 1.5 T scanner with round-the-clock monitoring; the figures are illustrative inputs chosen for this article, not a measurement of any specific product.
| Operating state | Average power (kW) | Hours per day | Energy per day (kWh) | Share of daily energy |
|---|---|---|---|---|
| Active state (scanning) | 31.4 | 7.5 | 235.5 | 59.9% |
| Idle state (ready-to-scan) | 14.6 | 3.5 | 51.1 | 13.0% |
| Stand-by mode (overnight) | 8.2 | 13.0 | 106.6 | 27.1% |
| Total | 16.4 (24 h average) | 24.0 | 393.2 | 100% |
Notice that stand-by mode runs for more than half the day yet delivers no diagnostic output at all, and still takes more than a quarter of the daily energy. Real hospital data show the same pattern, with the stand-by share often reaching 30 to 40 percent because of the continuously running cold head.
The magnet coils are wound from superconducting wire that only carries current without resistance near 4 kelvin. A cold head and a water chiller therefore run around the clock to recondense the helium that slowly boils off. Switching the cooling system off is not an option for a conventional magnet, so the floor under every MRI scanner is set by cryogenics, not by how busy the department is.
The idle state is the part you can actually influence. Many scanners stay in ready-to-scan between patients and during the lunch gap, drawing roughly twice the stand-by power. The active state depends on the protocol: diffusion-weighted imaging (DWI) and time-of-flight sequences draw noticeably more power than a simple T1 or T2 acquisition, and thin slices extend the acquisition time.
You can estimate yearly power consumption for any MRI unit with one sum: multiply the average power of each state by the hours spent in it, add the states together, then scale to the year.
$$E_{day} = \sum_{s} P_s \times t_s$$
$$E_{year} = E_{day} \times 365 \qquad \text{Cost}_{year} = E_{year} \times \text{tariff}$$
Here \(P_s\) is the average power in kW for state \(s\) and \(t_s\) is the hours per day in that state. Keep the units straight: kilowatt multiplied by hours gives kWh, the unit your utility bills in, and 1,000 kWh equals one MWh.
Using the hypothetical scanner above, the daily total is 235.5 + 51.1 + 106.6 = 393.2 kWh. Over 365 days that is 143,518 kWh, or about 143.5 MWh. At a tariff of $0.13 per kWh the yearly electricity bill is about $18,657; at $0.17 per kWh it rises to roughly $24,398. Compare that with the 100,000 kWh yearly benchmark often quoted in the literature and the result sits comfortably above it, which is typical for a busy 1.5 T system.
The sentence "a scanner uses about 16 kW on average" is also true: 393.2 kWh divided by 24 hours is 16.4 kW. Thinking in averages is useful for sizing the room's power distribution and cooling, while kWh is what drives the cost.
Large numbers like 143 MWh are hard to picture, so compare them with things you know. An average household in the United States uses on the order of 10,000 kWh per year, so the MRI scanner in the example consumes about as much electricity as fourteen homes. A typical refrigerator needs 1 to 2 kWh per day, and the example scanner uses that much energy in under six minutes of stand-by mode.
A single examination also has a footprint. At an average active power of 31.4 kW, a 45-minute brain scan costs about 23.6 kWh in the active state alone, and the same patient slot also carries a share of the idle and stand-by energy spent keeping the system ready.
Electricity is only half of the story, because helium is the other large operating input. A superconducting magnet sits in a bath of liquid helium, and a conventional system holds well over a thousand liters of it. Helium is a finite resource extracted alongside natural gas, and medical imaging is one of its largest single uses, which is why prices and supply have been volatile for years.
Shutting a magnet down deliberately is called a ramp down. A dedicated magnet power supply removes the current slowly, over hours, so the helium is preserved. A quench is the uncontrolled alternative: part of the coil stops superconducting, heat is released, and the cryogen boils off in seconds and vents outside the building. A quench can cost tens of thousands of dollars in lost helium plus extended downtime, which is why operators prefer to keep the magnet cold rather than switch it off.
Modern designs reduce the dependence. Helium recovery systems capture boil-off gas, sealed low-cryogen magnets hold only a small fraction of the usual inventory, and research into helium-free magnets aims to remove the issue altogether. These designs also tend to lower the stand-by load because less cooling work is needed.
Inside the equipment room, four subsystems account for almost all of the draw. The gradient amplifiers deliver the highest instantaneous power, with rating figures such as 2,000 V at several hundred amperes, and an RF amplifier in the 20 to 30 kW class feeds the transmit chain. Behind them sit the chiller and compressor for the magnet, the water cooling for the gradient, and the room's air conditioning. Together with the console computers they explain why the minimum power is well above zero.
The coil choice affects image quality much more than power: measurements on a phantom with a 16-channel and a 32-channel head coil showed no meaningful difference in consumption. The sequence parameters matter more, especially slice thickness and the number of averages, since they set the acquisition time at high gradient duty.
Hospitals increasingly report primary energy, not just metered electricity, because a kilowatt-hour from the grid hides the fuel used to generate it. Primary energy is the metered electricity multiplied by a national primary energy factor (PEF) that reflects the local energy mix. With a factor of 2.1, the 143.5 MWh yearly example becomes about 301 MWh of primary energy; with a cleaner grid and a factor near 1.5 it falls to roughly 215 MWh.
This matters for the carbon footprint because healthcare is responsible for a measurable share of national greenhouse gas emissions, and radiology is among the most energy-intensive departments in a hospital. Environmental product declarations (environmental product declarations, or EPDs) published by manufacturers estimate lifetime primary energy for a scanner, but on-site studies have found that real clinical use can differ from an EPD by up to about 40 percent in either direction, depending on the usage pattern and the national energy mix. That gap is the best argument for metering your own MRI systems instead of relying on a declaration.
Comparing a decade-old scanner with its successor on the same site shows what technological replacement can achieve. In published two-week measurements, the newer model used roughly 10 percent less energy per day, with most of the gain in the non-productive idle and stand-by states rather than in scanning itself. Over a 10-year horizon that translates into hundreds of MWh of primary energy saved, which is a strong sustainability benefit but, as the authors stress, not a reason on its own to replace a working scanner: clinical and diagnostic criteria must lead.
Suppose a department trims its idle power from 14.6 kW to 10.2 kW with an automatic ready-to-scan timeout and reduces stand-by from 8.2 kW to 7.4 kW through a firmware-controlled eco mode. The new daily total is 235.5 + 35.7 + 96.2 = 367.4 kWh, a saving of 25.8 kWh per day, or about 9,417 kWh per year. At $0.13 per kWh that is roughly $1,224, and in primary energy terms (factor 2.1) it is about 19.8 MWh.
A growing body of study work now measures energy consumption in radiology instead of estimating it. Researchers at university hospitals, including work associated with Stanford University and other academic centres, have logged daily consumption for individual machines, and the broad findings agree: the operational energy of magnetic resonance imaging is dominated by the always-on cooling, and the energy-saving potential lies in the non-productive hours.
Some papers express the average rate in kW/h, which is technically a rate of change of power, whereas the clean unit for a rate of energy use is kW. When you compare studies, convert everything to kW and kWh first, and check whether the authors quote daily consumption for one scanner or the total for a whole department. Healthcare managers who skip that step end up comparing a single machine with a whole imaging floor.
The reason is physical. Superconductivity only holds at a very low temperature, around 4 kelvin, so the cooling load is a requirement of the technology, not a choice. The equipment that provides it is the chiller, the compressor and the cold head, and all three run whenever the magnet is energised. Newer technological replacement options lower that floor through better insulation and smaller helium inventories.
For a hospital with a sustainability plan, the energy consumption of MRI equipment is a fixed, measurable item. The environmental case rests on three numbers: the kWh metered each year, the primary energy factor of the local grid, and the helium inventory. A department that lowers its energy consumption by ten percent reduces its environmental impact, its bill and its exposure to healthcare carbon reporting rules at the same time. Next, look at car stereo power consumption.
Over 10 years the arithmetic becomes persuasive. The 9,417 kWh yearly saving from the earlier scenario becomes 94,170 kWh, and the primary energy at a factor of 2.1 is about 198 MWh. Even without counting helium, that is an environmental gain worth documenting. Every manufacturer now publishes some form of energy declaration, but only on-site metering tells you how your operational pattern performs, so your energy consumption data should sit beside the declaration in the purchase file.
In practice, a radiology team can track metered kWh per scanner each quarter, split by stand-by, idle and active state, and set a low energy target for the radiology department. Tracking the energy consumption of each unit this way shows which energy savings and energy efficiency measures actually work, keeps the operational data tied to clinical scheduling, and gives the radiology service a measured baseline for the energy consumption it reports.
You cannot switch a superconducting scanner off, but you can capture most energy savings with a short, ordered checklist. Start with the largest and cheapest items first:
Together these steps tend to produce energy efficiency gains of 10 to 20 percent without any loss of image quality, which makes them an easy case for a hospital's sustainability plan.
Per hour of operation a CT scanner is usually lower in average power, but an MRI's always-on cooling means its daily total is much less sensitive to utilisation. Stand-by is a large share of an MRI's energy consumption and a comparatively small share of a CT's.
The imaging chain is, but the magnet's cooling is not. Overnight the scanner sits in stand-by mode, drawing the cold head and chiller load that keeps the helium liquid.
A full shutdown means ramping the magnet down, losing or recovering helium, then re-cooling and re-ramping it, which takes days and is expensive. Operators keep the magnet cold instead.
Clamp a power analyzer on the main feed for at least two weeks, log in kW at one-second to one-minute intervals, then integrate the curve to get kWh and split it by state using the scan log.
Reliable figures come from metering, not from nameplate ratings. The usual method is a power quality analyzer clamped onto the scanner's electrical cabinet or power distribution unit, logging current and voltage at intervals of one second to one minute for days or weeks. The log gives a power curve in kW against time, and the area under that curve is the energy in kWh. Matching the curve against the scan log from the picture archiving and communication system (PACS) lets you assign every minute to the stand-by, idle or active state.
A peak demand rating tells an electrician how big the supply must be; it says almost nothing about yearly energy. A scanner rated for 60 kW of peak power may average only a quarter of that across a day, because the peak occurs for seconds inside a gradient-heavy sequence. Use the rating to size cables and breakers, and use the metered average to budget.
Short logging windows can mislead when you are budgeting MRI scanner energy. A single weekday sample overstates the active state, while a window that includes a weekend overstates stand-by mode. Two full weeks, as used in several on-site studies, smooths most of the variation in the metered kWh, and a full year captures seasonal chiller loads.
Higher field strength raises the demand on the gradient and RF chain, but the picture is subtler than "more tesla, more kilowatts." A 3T scanner with a wide bore and a high-performance gradient can use two to three times more energy in a long examination than a compact 1.5 T system, yet it may finish some protocols faster, and the stand-by floor is set mainly by the magnet's cryogenic design rather than its field.
| Factor | Effect on energy consumption | Controllable after purchase? |
|---|---|---|
| Field strength and bore design | Sets gradient and RF power class | No |
| Stand-by cooling load | Sets the daily floor, around the clock | No |
| Protocol and sequence choice | Changes active power and duration | Yes |
| Idle timeouts and scheduling | Changes hours spent in the idle state | Yes |
| Room cooling and chiller | Adds facility load to the total | Partly |
The practical lesson: field strength matters at the moment of purchase, while scheduling and protocol discipline matter every day afterwards.
Because the stand-by floor is fixed, yearly cost responds to the electricity tariff much more than to patient volume. The next table applies the example scanner's 393.2 kWh per day (143,518 kWh per year) to several tariffs, then shows what happens if the active hours fall by two hours per day. You can also check how many watts does an e scooter use.
| Tariff ($/kWh) | Yearly cost at 7.5 active hours | Yearly cost at 5.5 active hours |
|---|---|---|
| 0.09 | $12,917 | $9,787 |
| 0.13 | $18,657 | $14,138 |
| 0.17 | $24,398 | $18,488 |
| 0.22 | $31,574 | $23,926 |
Cutting scanning time by 27 percent (7.5 to 5.5 hours) trims the cost by only about 24 percent, because the stand-by and idle blocks stay put and in practice the unused hours simply move into idle. That is the reason experienced energy managers focus on idle timeouts before they talk about patient throughput.
For a hospital with several MRI machines, a $20,000 line item per scanner is small next to the purchase price or the service contract, yet it is permanent, it scales with tariffs, and it is directly linked to the institution's carbon target. That is why energy teams treat the MRI scanner as a target worth metering even when its share of the hospital's bill is modest.
Within a hospital, radiology consistently ranks among the top departments for electricity, and the MRI suite usually leads inside radiology, ahead of CT and ultrasound. Because healthcare produces a notable share of national greenhouse gas emissions, every kWh trimmed from an MRI scanner counts toward climate change targets, and that link is what drives interest in green radiology.
Initiatives usually follow a similar order: measure, switch off what is not needed, tune protocols, then buy better. The same steps apply whether you operate one scanner in a clinic or a dozen in a hospital network.
Asking these four questions about the imaging equipment before signing a contract usually reveals more about lifetime energy consumption than any brochure headline figure, and it gives you a baseline for judging future energy consumption against what the vendor promised, including the cooling equipment.