Blind Spot Austen Blake Josh Kyle Tyler Overview

Blind Spot Austen, Blake, Josh, Kyle, Tyler

Overview • Nuclear energy history • How nuclear energy works at TMI • Timeline of events at TMI • Impact of lessons learned at TMI

Leo Szilard • Motivation: Read an article in 1933 written by Ernest Rutherford that rejected the idea of using atomic energy for practical purposes. • Filed for a patent in 1936 and assigned it to the British Admiralty to ensure secrecy. • Accepted an offer in 1938 to conduct further research at Columbia U, where he was joined by Enrico Fermi. • Concluded uranium was the element capable of sustaining a chain reaction.

Uranium Chain Reaction Weak nucleus broken by other neutrons When these atoms break up they release showers of more neutrons which in turn break up more atoms, etc. (see side picture) Dangerous because of the chance of a runaway reaction. The insertion of boron-free graphite regulated the number of loose neutrons by absorbing them easily Successfully controlled by humans on December 2, 1942.

Impact of War • With the successful engineering of the two atomic bombs used against Japan in WWII, there were new ideas stemming out aiming to take advantage of the use of chain reactions: Cars running on tiny uranium pills that were powerful enough to heat an entire house throughout winter. o Transmuting base metals to gold so cheaply that pipes used in operating systems might be made of gold to reduce maintenance costs. o Fluoroscopes in shoe stores that allowed customers to see a full-motion X ray of how their new shoes fit when they wriggled their toes o And. . . o

Nuclear Power Plants • Pushed into production because the military needed them for propulsion on board long range ships. • The first commercial nuclear power plant began running in 1957 in Shippingport, PA. • By 1966, the country had 15 nuclear power plants running, 9 under construction and 22 on order. • General Public Utilities (GPU) owned 3 big utilities and would eventually erect Three Mile Island Unit 1 in 1974 and Unit 2 in 1978. Unit 2 would soon be known as the site of the worst civilian nuclear accident in US history.

Turning reaction into energy Controlled uranium reaction produces heat Water added to the heat produced steam Steam powered turbines to create 880 megawatts Three stages of pipes(chain links) to accomplish this: o Reactor coolant (Blind Spot) o Steam making pipes o External Cooling

Turning reaction into energy cont. . . Reactor Coolant Pipes • Reactor core to steam • Extract heat using high temp, high pressure water • Entirely inside containment building • Invisible to employees, relies completely on sensors Steam Making Pipes • Reactor coolant pipes through turbines • Ultrapure water flashed into steam • Located in turbine building • Visible, reachable to employees

Turning reaction into energy cont. . . External Cooling • Turbines to environment • Extracted waste heat • Large cooling towers • Visible to entire city

Early Hours of March 28 th, 1979 on Three Mile Island • Unit 1 power plant was down for repairs and testing • Unit 2 was running at 3 million horsepower, enough to supply a city of five hundred thousand people • In Unit 2, Don Miller and Harold Farst opened up a part of the steam-making loop for maintenance. (The system had valves to isolate this area)

Maintenance in Unit 2 • The men spent all night trying to shake loose tons of tiny palstic beads that were causing a jam in a demineralizer tank. • During the maintenance a few ounces of water seeped back into the compressed air lines • At 36 seconds past 4: 00 a. m. , the leaking water reached the control line to the big valves controlling the condensate polishers. • The automatic controls interpreted this tiny bit of water in their air lines as deviation from proper conditions and so shut all the valves that let coolant through. • This caused a roadblock effect. • The inertia of five thousand gallons of water a minute tore on of the big pipes loose in the turbine building, pulled out controls, and sprayed the place with scalding water.

Maintenance in Unit 2 • With the steam-making pipes shut off, the water in the reactor coolant loop had no place to dump its heat. • The reactor coolant loop had only one open space for the water to expand into. It was the pressurizer tank. • The pressurizer tank acted like a shock absorber to the reactor coolant piping. • Automatic controls maintained the right balance of water and steam by either cooling or heating the contents of the pressurizer tank. • At the top of the pressurizer was a safety valve to let off steam if pressure rose too fast for automatic controls. The pilot operated relief valve (PORV) opened as intended, reducing pressure by letting steam out the top of the pressurizer when the water level rose.

The Rubble Maker • That morning the pressur stabilized but when the electronic command came to close the PORV, the valve stuck open. • A warning horn went off shortly after 4: 00 a. m. , warning that pumps in the steam-making pipes had shut down. • Three emergency pumps were coming on-line automatically, to keep water moving along in the stem-making pipes and assist the flow of heat out of the reactor. • An indicator showed that an electric signal ordered the PORV to open and vent any pressure over 2, 255 pounds per square inch. • Shortly after the pressure dropped and the PORV indicator went out. • Operators thought this meant PORV had closed, but it just meant the command had been sent, nothing showed that it was actually still stuck open

The Rubble Maker • During the first minute, the level of water showing in the pressure tank was dropping. • An operator turned on high-pressure pumps to throw some extra water into the reacotr coolant pipes. • The pressurizer water level began going up from its low point of 158 inches. • The highest possible reading was 400 inches on the gauge. • The rising water did not stop, after four minutes the water level hit 300 inches. • Pressure was dropping, but the water level kept going up. • Past 400 inches on the gauge, the operators knew there would be nowhere for the rising water volume to go.

The Rubble Maker • There was now a fear that the pipes may "go solid", meaning the pipes fill with water leaving no air space at all. • There was not an instrument or gauge showing how much water actually sat in the reactor vessel, or in the piping, or in the steam generator. Only the pressurizer tank had a water level indicator. • The PORV on the pressurizer was letting steam out at the rate of 220 gallons of water per minute. It was this leak at the top of the pressurizer that sent the water level so high, fooling the operators.

Missed Warning • September 1997 (~1. 5 years earlier) the Davis Besse reactor had a similar problem • Differences: 1. They were running at low power 2. The issue was discovered after 20 minutes • Why did they not share? They thought this problem was unique to them. • They should've warned others about the relief valve mislead them into thinking the coolant was going solid.

Who's to blame? • With no warning, the crew guarded against the TMI going solid • They cut back the flow of high pressure cooling from 500 gallons per second to 25 gallons per second. • The control room designers and management may be to blame here. o A memo was sent out a year before stating that the control room would cause problems some day. o nothing was done

Problems Continue • Water level on the pressurizer gauge continued to rise • (4: 06 AM) water stopped climbing as it reached the top of the gauge (400 inches) • The water would creep down while they let more water out of the primary system, then creep back up to 400 inches • Errors were being recorded from 100's of alarms but the printer could only handle 15 lines per minute. It was soon 2 hours behind.

Problems Grow • (4: 20 AM) The pressurized water level was hovering around 370 inches which made time to straighten out the mess at the condensate polishers • (5: 00 AM) Returned to the control to find the four giant pumps that forced water through the reactor coolant pipes were shaking and falling apart. The pump impellers met cavities of steam, sped up, then slammed into solid water and slowed down. • This caused the pipes to crack

Call for Help • After pipes cracked they called Brian Mehler in to help • (6: 00 AM) Mehler arrived in the control room and received a short briefing • For a few minutes, Mehler pondered how the water pressure could be down if the pressurizers water level was up

A New Set of Eyes • Mehler had 2 theories: 1. Blown circuit breaker took out heaters in the pressurizer tank. These normally come on when pressure drops 2. There is a small leak in the coolant system • He knew that pressure was up in the containment building and it was getting warm in there which indicated a steam leak. • After looking up the PORV temp he noticed it was too hot

First Good Thing • He then asked if he could close a valve that would isolate the pressurizer tank for the PORV • This caused the primary coolant pressure to head back up • Finally the overheating stopped

Legacy of Three Mile Island • Economic and environmental effects o Clean up cost estimated at $1 billion and takes fourteen years o Tests show no long term health effects • Reform for the nuclear industry o Redesign of control room layout o New guidelines on safety, training, and emergency preparedness inspection,

The Blind Spot • Inner workings are too complex or hidden form operators • Operators unaware of how little of information they are receiving • Hidden Instruments • Misreading or misinterpreting instruments and data

Avoiding Blind Spots • Systems with tight coupling and high interactive complexity should be avoided o Time-driven event where one event leads to another in short order o Systems subject to chains of unexpected failures o Keep things simple and well organized • Creating a high reliability organization o Make safety the main priority o Create deep redundancy in systems o Keeping sharp with practice and emergency drills o Learn form trial and error

Summary • Nuclear energy history • How nuclear energy works • Timeline of events at TMI • Impact of lessons learned at TMI

Questions/Discussion