5 Unit Conversion Mistakes That Cost Real Money

In September 1999, a $125 million spacecraft reached Mars and promptly disintegrated. The Mars Climate Orbiter had traveled 416 million miles over nine months. Everything worked except for one detail: the software that calculated thruster output used pound-force-seconds while NASA's navigation team expected newton-seconds. That factor-of-4.45 mismatch sent the probe 57 km into the Martian atmosphere instead of the planned 226 km orbit.

Unit conversion errors are not just math class problems. They destroy spacecraft, strand airplanes, sink warships, and put patients in hospitals. Here are five documented cases where getting the units wrong carried a real price tag.

1. NASA's Mars Climate Orbiter ($125 Million, 1999)

Lockheed Martin wrote the ground software that tracked thruster firings. Their code output impulse data in pound-force-seconds. At the Jet Propulsion Laboratory in Pasadena, the navigation team imported that data into trajectory models calibrated for newton-seconds, the SI standard. Nobody caught the discrepancy during the 286-day cruise to Mars.

One pound-force equals 4.45 newtons. Every course correction over those nine months compounded the error. The investigation board called it "the root cause" and noted that existing verification procedures should have flagged the mismatch months earlier.

The spacecraft itself cost $125 million; the total mission cost, including development and operations, reached $328 million. Verifying that both teams were working in the same impulse units — pound-force-seconds versus newton-seconds — takes two seconds with a checklist. The absence of that step cost NASA a spacecraft and years of planned science.

2. Air Canada's Gimli Glider (Near-Catastrophe, 1983)

On July 23, 1983, Air Canada Flight 143 ran out of fuel at 41,000 feet over Red Lake, Ontario. Both engines on the Boeing 767 went silent. Captain Robert Pearson glided the unpowered aircraft 100 km to a decommissioned air force base in Gimli, Manitoba. All 69 people on board survived.

The 767 was one of Air Canada's first metric aircraft. Its fuel gauges were inoperative that day, so the crew calculated the load manually. They needed 22,300 kilograms of jet fuel for the Montreal-to-Edmonton route.

The crew used a density factor of 1.77, correct for pounds per liter, the standard on every other aircraft in the fleet. But the 767 required kilograms per liter: 0.803. They measured 7,682 liters in the tanks, multiplied by 1.77, got 13,597, and treated that number as kilograms. They ordered 4,917 liters more. The aircraft departed with roughly 10,100 kg, less than half of what the route required.

The conversion between pounds and kilograms is well-documented. The problem was assumption, not difficulty: the crew applied a familiar number from muscle memory without checking which unit system the new aircraft used.

3. The Vasa Warship (National Embarrassment, 1628)

The Vasa was a 64-gun Swedish warship that sank on her maiden voyage, August 10, 1628. She made it about 1,300 meters out of Stockholm harbor before a gust of wind heeled her to port. Water poured through the open gun ports. She went down in front of thousands of spectators.

When archaeologists raised the Vasa in 1961, they found that the rulers used by the shipbuilders did not match. Workmen on one side of the hull had used Swedish feet (twelve inches). The other side used Amsterdam feet (eleven inches). The ship was asymmetric, heavier to port.

The Vasa was also top-heavy by design, with too many guns stacked high. But inconsistent measurements between two construction teams made the instability worse. Two groups building one ship, each assuming their "foot" was the same length, produced a vessel that could not survive its first harbor crossing.

4. Medication Dosing Errors (Ongoing, Billions Per Year)

The examples above are dramatic, one-time failures. Medication dosing errors happen every day, and their cumulative toll dwarfs any spacecraft.

The Institute of Medicine estimated that at least 1.5 million preventable medication errors harm patients in the United States each year. The gap between milligrams (mg) and micrograms (mcg) is a factor of 1,000. A pharmacist selecting "mg" instead of "mcg" from a dropdown menu delivers a dose one thousand times larger than intended.

One documented case involved a pharmacy compounding nutrition for a neonate. A pharmacist entered the zinc dose in milligrams instead of micrograms. The error passed through six staff members uncaught. The infant received a fatal overdose. The two units were adjacent items in a pull-down menu.

Temperature conversions matter in healthcare too. A body temperature of 39.5 C is 103.1 F, a serious fever. You can check that with our Celsius to Fahrenheit converter. Misreading a thermometer's scale or failing to convert when transferring records between systems can delay treatment decisions.

5. Columbus and the Wrong Size of Earth (1492)

Columbus did not discover that the Earth was round. Educated Europeans already knew that. What he got wrong was how big it was.

He relied on Al-Farghani, a 9th-century Persian astronomer who measured one degree of latitude at 56.67 miles. That number was accurate. The problem: Al-Farghani meant Arabic miles, each about 1,900 meters. Columbus assumed Roman miles, each about 1,480 meters. That misidentification shrank the Earth by roughly 25%.

His final estimate put the distance from the Canary Islands to Japan at 3,080 nautical miles. The actual distance is over 10,000. Had the Americas not been in the way, Columbus and his crew would have died of thirst somewhere in the Pacific.

Converting between distance units is trivial now. A quick check of kilometers to miles takes no effort. In 1492, the problem was not laziness but the absence of standardized units across cultures. Every region had its own "mile."

What These Failures Have in Common

These stories span 500 years and fields as different as space exploration and pharmacy. Three patterns repeat:

Handoffs between teams. The Mars Orbiter failed at the boundary between Lockheed Martin and JPL. The Gimli Glider failed where Air Canada's old fleet met its new metric aircraft. The Vasa failed between two groups of shipwrights. Whenever data crosses a boundary, units need explicit confirmation.

Familiarity breeds assumptions. The Air Canada crew had calculated fuel loads hundreds of times using pounds. The 767 was their first metric aircraft. Habit overrode procedure.

Small differences compound. Columbus's unit error was 25%. The Mars Orbiter's was 4.45x. Neither seemed dramatic on any single data point. Over months or thousands of miles, the errors became catastrophic.

How to Protect Yourself

You probably are not navigating to Mars. But you might be converting a recipe, reading a prescription label, or comparing fuel prices in Canada.

1. Label everything. "22,300 kg" is unambiguous. "22,300" is a guess.
2. Verify at handoffs. When a number comes from someone else, confirm the unit before using it.
3. Use a converter. Whether you need liters to gallons for fuel, Fahrenheit to Celsius for weather, or kilometers to miles for a road trip, five seconds of conversion beats guessing wrong.
4. Sanity-check the result. A 767 needing only 4,917 liters for a transcontinental flight should have raised eyebrows. It did not.

The Mars Climate Orbiter is gone. The Gimli Glider survived on luck and pilot skill. You can do better than both by checking the units first.

Sources: NASA Mars Climate Orbiter Mishap Investigation Board Phase I Report (1999), Transportation Safety Board of Canada Report (1985), Vasamuseet (Stockholm), Institute for Safe Medication Practices (ISMP), IEEE Spectrum, US Metric Association