20 kWh electric car battery boils off less than 28 liters of water
TNT: 4.6 kJ/g (lousy fuel but does nice boom) (boils 1.77g/g)
gasoline (with oxygen): 13 kJ/g (boils 5g/g)
reactor-grade uranium (4%) in light-water reactor: 3,456,000 kJ/g (boils 1,332,000 g/g == 1.3 tons of water per gram of U, 266,000 times more energy per mass than gasoline)
natural uranium in fast breeder reactor: 86,000,000 kJ/g (boils 33,153,000 g/g == over 33 tons of water per gram U, 6.6 million times more energy per mass than gasoline)
assuming fuel consumption of 5l/100km, or 20 meters per mL, assuming 0.75 g/ml == 26.7 m/g, we get 7090 km per gram of enriched uranium in LWR, or 176,000 km per gram in breeder
for a T72 tank, with 263l/100km, we get range of 134 km/gram or 3346 km/gram.
hint on civilian accidents (Goiania - 93 grams of caesium-137 chloride radiotherapy source, abandoned in a hospital - lawsuit prevented removal from site)
multiplication coefficient - k<1 for dieout, k>1 for exponential growth
criticality - shape vs mass, material type - metal vs solution - criticality accidents
neutron economy, reactivity
prompt neutrons (within 1e-14 sec, most) vs delayed neutrons (msec to minutes past fission, up to 3%)
delayed is key to stable reactor control, prompt-critical when reactivity enough to not need them
prompt critical vs controlled k=1
stable reactor slightly subcritical related to prompt neutrons, delayed ones keep pushing it forward - two populations
Pu gives fewer delayed neutrons, control at high Pu proportion in MOX challenging
self-control in accidents, pulsed mode, negative feedback (thermal expansion) - cricitality accident in water, pulses
weapons
brief principle
uranium vs plutonium
gun vs implosion
plutonium production
fission/boost/fusion, promise of another lecture
uranium enrichment
uranium hexafluoride
isotope separation - any method where lazier behavior of heavier nuclei can be used
calutrons
diffusion
centrifuges
Stuxnet worm, cyberwarfare (separate lecture?)
laser - AVLIS, MLIS - slight difference of absorption spectra used for selective ionization of isotope
high cost
uranium vs plutonium as fissile material, production vs enrichment
fissile breeding
spent fuel reprocessing
graphite-moderated reactors
heavy water moderated reactors
fuel burnup degree - 239Pu-240Pu
fuel-grade, weapon-grade, supergrade Pu - 240Pu spontaneous fission, decay to 240Am, energetic gamma emissions of 240Am (Pu mostly just alpha)
Manhattan Project - spontaneous fission - Thin Man to Fat Man
fast reactors
no moderator
high enrichment
produces own fuel, solution to uranium shortages - energy extraction from uranium 60% vs 1%, can feed slow reactors as mixed-oxide (MOX) fuels, uranium price drop halts deployment
less transurans in spent fuel
first plutonium breeder reactors
Hanford
Windscale
RBMK reactors, Magnox - dual-use, can both breed and give energy, RBMK can replace fuel during operation
graphite as moderator
25 Jan 1939 first US fission test, multiplication detected by irradiating uranium sample with Ra-Be source
Fermi estimates 1.73 neutrons per fission, real value about 2.4 per U-235 atom
first piles from 10x10x30 cm graphite blocks, 20x20x20 cm uranium cans, Ra-Be neutron source below, measuring multiplication
Feb 1940 first tests by Enrico Fermi, but results discouraging - unknown impurity was eating neutrons
arc lamp specialist consulted
ordinary graphite too high in boron; thermal purification development took two years, resulting in nuclear graphite
boron control in production very tight, down to borax soap prohibited for workers' garment washing
heavy water was a backup solution if this was not solved
Siemens Plania graphite too high in boron too, Germans couldn't solve, opted for heavy water instead
first reactors
1942, competing designs for reactor production, decision deadline forced by Groves at Oct 5, Met Lab got a week, time more important than money
compromise: intermediate reactor with helium cooling built in Argonne, run until June 1943, then dismantled and Pu extracted
full-scale facility, "Mae West" design, built in Hanford, operating by March 1944, producing 100g Pu per day
Chicago Pile 1
under a sports stadium viewing stands of University of Chicago, built by Metallurgy Lab, run by Enrico Fermi
"a crude pile of black bricks and wooden timbers" according to Fermi
worries about structure disintegration under neutron bombardment - brick fusing, disintegration - first reactors had to be built without knowledge, cyclotrons too weak to test
dissolved, oxidized, precipitated again - Pu now in solution with further 4-orders of magnitude radiation reduction by removal of fission products, 1/100,000 of original activity
lanthanum salt added, hydrogen fluoride added, lanthanum fluoride precipitated, lanthanides coprecipitated - bismuth phosphate won't remove these
oxalic acid added, plutonium reduced back to Pu(IV)
lanthanum precipitation repeated with added potassium hydroxide
liquid removed, precipitate redissolved in nitric acid, forming plutonium nitrate
1200 liters of original solution now turned into 30 liters
gen4 - in development - sodium fast reactor, molten salt reactor, lead cooled, supercritical water, gas cooled fast; liquid metals and salts allow operation on atmospheric pressure
gen5 - theoretically possible, not considered; gas-core, fission-fragment, hybrid fusion-fission (fusion neutrons for fissioning - subcritical reactor)
basic reactor classes
graphite
Magnox, gas-cooled
RBMK, water-cooled
heavy water
CANDU
water
boiling water, high pressure needed - direct steam-to-turbine
pressurized water, even higher pressure needed - heat exchanger for steam generation
fast, metal-cooled (NaK, Pb-Bi)
molten salt - coolant can be fuel, no pressure
reactor characteristics
void coefficient
positive vs negative
boiling reactors
temperature coefficient, Doppler coefficient
TRIGA, extremely high negative coeff, pulsed operation
startup, role of neutron sources in core
hydrogen production - water radiolysis, water-cladding (zirconium) reaction
xenon poisoning, iodine pit
iodine-135 is one of fission products, half-life 6.57 hours, decays to xenon-135
xenon-135 is the best neutron absorber ever
either decays spontaneously, with 9.14 hour half-life, to caesium-135
situation can be just waited out, this problem self-heals (not an option if the test MUST be done or if we're a submarine)
or can be burned, as captured neutron transforms it to
equilibrium reached after a while of operation at steady state
when power lowered, xenon level increases and further lowers power
when power increased, neutron flux burns more xenon and power increases further - positive feedback loop in both cases
reactor can be put out of commission for a while, made non-restartable until xenon decays
can be steamrolled over if there's enough reactivity on hand - usually high-enough enrichment, with high content of burnable poison to compensate
doomed Thresher submarine; submarine reactors need highly enriched fuel
anatomy of power plant
heat source
thermal vs electrical reactor power - MWt vs MWe
steam source
turbine
generator
rotating magnetic field as electron pump
rotor magnetic field generated by electromagnets, powered by magnetization generator on the same shaft
generator must be kept in-phase; attempt to push it forward increases torque and output power
too slow turbine rpm, generator starts "motoring" - taking power from the grid and spinning actively, undesirable
hydrogen cooling/lubrication - good coolant, MUCH lower windage losses - same reason as helium in hard drives
Chernobyl plant
4 blocks, pairs of reactors sharing some infrastructure and stack
filled with helium at 0.5 MPa to aid with heat transfer through pellet-cladding gap
pellets in retaining rings, held axially with springs, each rod has 3.5kg of pellets
max temperature 600 (700?) °C, then hydrogen slowly starts being produced and embrittles the alloy
avg power 15.3 kW/m, peak 35 kW/m (20.5/42.5 kW/m for RBMK1500)
(image illustrative only)
assembly
two subassemblies, each with 18 tubes, around 1.3cm thick carrier rod
total length 10.015 meters
instrumented assemblies have central 1.5cm tube, 2.5mm thick wall, neutron flux detectors inside
tubes held in 10 stainless steel plates, 36cm apart
subassemblies joined with a cylinder - lower neutron flux in central plane of the reactor
exposed to water/20% steam at 7 MPa, free oxygen present, risk of nodular/crevice corrosion
114.7kg (111.2?) of uranium per assembly, 20 MW.day/kg burnup (480 MWh/kg, 55000 MWh/assembly)
192 tons of fuel in reactor (92.16 terawatt-hour)
fuel and channel failures
at 650'c the zircaloy starts plastically deforming, fuel channel walls deform
at 900'c the fuel cladding fails due to ballooning via internal pressure vs depressurization of the outside
at about 1000'c zirconium reacts with steam, produces hydrogen, until pressure drops to atmospheric and steam is not available anymore
here the operator should stop feeding coolant
at 1450'c, stainless steel grids melt
at 1930..2050'c control rod alumina cladding melts, at 2330'c boron carbide melts, at 2600'c ceramic formation begins
fuel replacement
remote-controlled gantry crane
position above channel, mate, equalize pressure, pull out assembly, pull in fresh assembly, move assembly to cooling pond
fuel can be replaced in a running reactor
better reactor uptime
easier to do lower burnup, for better-quality weapon-grade plutonium production
choose between longer in-reactor stay for better extraction of energy, shorter stay for better plutonium with less Pu-240
impossible to find by plant observation which assemblies are replaced, impossible to infer level of Pu production - observable reactor shutdowns in too short intervals can hint
at nominal power level two assemblies per day are replaced, five per day at peak
large gantry crane requires tall reactor building of rectangular shape, makes containment impractically large
spent fuel
U-235: 4.5 kg/t, 0.45%
U-236: 2.4 kg/t, 0.24%
Pu-239: 2.6 kg/t, 0.26%
Pu-240: 1.8 kg/t, 0.18%
Pu-241: 0.5 kg/t, 0.05%
startup neutron sources
12 rods acting as neutron sources to aid with startup - see Fermi's source under first subcritical piles to measure multiplication factors
the provided neutron flux is multiplied by the reactor's "k"; rise of power much smoother, less abrupt
primary sources for fresh reactor startup - conventional Cf-252 (spontaneous fission), Pu-238/Am-241/Po-210/Ra-226 with beryllium
sensitive to neutrons which burn up the isotope
Pu-Be can be affixed to control rods and removed from core when running, or clad with cadmium foil which is transparent to fast neutrons produced by source and opaque to thermal ones
secondary sources, inert at first, activated in the reactor
Sb-Be - antimony becomes activated, powerful gamma photons kick out neutrons from beryllium-9 - photoneutrons
fuel itself can provide enough, except if fresh or after extended shutdown
control rods
made from boron carbide (enriched with boron-10), with added dysprosium titanate
0.97m long segments
24 SAR, shortened absorber rods - axial control
3 elements, 3050 mm, inserted from below
5-element graphite displacers on top
24 AC, automatic control - 5 elements, 5120 mm; in 2 groups
no displacers
12 local auto control rods
12 average power control rods, in 3 banks of 4 rods each
139 MR, manual radial distribution control, 5 elements
having a 1.5m telescopic spacer, then 5-element (4.5m long) graphite displacer
of them, 24 local emergency protection (LEP), 2 rods per zone
leaves 115 manual control rods
24 ER, emergency rods, 5 elements
same design as MR
graphite displacers:
graphite displaces water, increases reactivity
graphite > water > control rod
when rod fully retracted, graphite section is in the middle of the core, augments neutron flux there (see fuel assembly center part)
independent cooling circuit, rods kept at 40-70'C
rod has to pass through water, narrow clearance between channel wall and rod - high resistance
therefore slow insertion - 40cm/sec, full insertion takes 18-21 seconds
after Chernobyl the servos on other reactors were replaced with stronger ones, shorter insertion times; graphite tips also replaced
after Chernobyl channels redesigned, rods move in gas, cooled by thin film of flowing coolant
local criticality possible if no absorber in a group of 15-20 fuel elements
high positive reactivity with few additional rods in core lowers power stabilization period to 3 minutes only
in case of excursions the period can be as low as few seconds
rod suspended on steel cable (later steel tape) wound on a drum, driven by DC motor, with selsyn indicator and endstop switches
electromagnetic braking used to slow down rod descent under gravity to avoid damage
post-Chernobyl 24 rods get gas-filled, water-film-cooled channels, used for scram
displacer effects
4.5m of graphite in the middle of reactor when rod fully retracted
1.5m of water above and below
insertion of rod displaces water, locally increases reactivity - "end-rods effect"
end-rods effect was not fully understood, occurs only in some cases of neutron distributions in core
chief-designer organization assumed it can happen only with neutron field disturbed downwards, was wrong
additional absorbing rods
240 static boron-based absorbers placed in core when all fuel is fresh, to compensate for excess reactivity
gradually removed and kept empty or replaced with fuel during operation
165 tons of uranium in fresh reactor, increased to 192 tons when stationary operation reached
with fewer additional absorbers, void coefficient increases
possible to keep them in permanently, and maintain low void coefficient, by using more enriched fuel
fuel originally at 1.8% enrichment, then increased to 2%, then after Chernobyl to 2.4% (which allows 80 permanent absorbers in core)
for RBMK1500 in Ignalina: with 2.4% enrichment added 0.41% of natural-isotopes erbium(III) oxide into fuel matrix as burnable poison
allows removal of permanent absorbers, higher burnup
core layout
blue startup source
green control rod
red automatic control rod
yellow shortened from-bottom control rods
grey pressure tubes (fuel channels)
coolant system
core divided to two halves, each independent; each half has two horizontal drum steam-water separators
risers from the bottoms of the drums feed the main pump manifold
tops of the drums feed high pressure steam to a pair of turbogenerators
return water from turbine condensers is fed to the drums
a nightmare of pipes
separator drums
horizontal cylinders, two on each side
2.3m dia, 30.7m long, 335.6 m3 (bit less, this volume is for Ignalina), top pressure 7.5 MPa
each has
432 inlet fittings, 70mm dia, in four rows (2 rows on each side)
inlets face baffle plates that dissipate coolant kinetic energy
14 steam outlet fittings, 300mm dia
12 downcomers, dia 300mm
1 feedwater inlet, dia 400mm
distributed in-drum to each downcomer, mixed in their inlet, in heat shroud to avoid thermal loads on drum/pipes
5 steam connectors between adjanced drums, dia 300mm
8 fittings for level meter connection, dia 50mm
14 fittings for water and steam sampling, pressure measurement, dia 10mm
fed with water-steam mixture from reactor, gravity-separated in drums
level of water in drums is critical
two levels of reactor trip, -600mm and -1100mm, depending on power level
level drop in accident conditions indicates absence of feedwater supply
vented via pressure relief valves
these commonly activate at shutdown of turbine or a transient, make noise, employees used to this
piping circuit
downcomers from both drums on each side come to a suction header
horizontal pipe 21.074 m long, 1.02 m outer dia, 13.4 m3
common suction header feeds four pumps
pumps feed common pressure header
horizontal pipe 18.204 m long, 1.04 m outer dia, 11.8 m3, parallel to suction header
bypass line between pressure and suction headers - six lines with control and check valve
used for gravity/convection flow when main pumps not operating
pressure header feeds 20 group distribution headers, total 32.6 m3
outer dia 325 mm each
each group distribution header feeds 41-43 fuel channels
each bottom pipe to channel has its own control valve and flow meter, inner dia 50 mm
each channel has volume of 78.6 dm3
CPS cooling circuit
CPS top storage tank above reactor
feeds radial reflector cooling channels
feeds control rod cooling channels, drains to bottom CPS storage tank
independent on main cooling system, usable for emergency cooling
turbine circuit
steam from drums collected in steam lines
steam lines interconnected, fed to SDV-A, SDV-C, MSV I-III, turbines
steam line split; part comes to high-pressure turbine stage
second part goes to a heat exchanger that reheats steam from high pressure turbine stage
reheated steam goes to two low-pressure stages
from low-press stages steam-water goes to condensers, cooled with water
water from condensers is preheated, filtered, fed by 7 main feedwater pumps through mixers to drum separators
also 6 auxiliary pumps, used to pump main coolant circuit full during startup, to maintain pressure when feedwater pumps fail
aux pumps also used for startup, shutdown, and low power operation
ventilation for emergency or post-shutdown cooling
both compartments have five blowout panels venting at 2 kPa, to accident steam release shaft
group distribution compartments: group distribution headers, lower water piping
pump compartment: main circulation pumps, suction and pressure header
coolant system temperatures
inlet 265-270'c, usual 270'c/8.2 MPa; saturation temperature at 8.2 MPa=284'c
boiling starts at about 2.5m from inlet to core
controlled by inlet temperature, pressure gradient, water flow, reactor power
individual valves for each fuel channel
complex operation, stuck valve caused channel damage on Block 1 in 1982(?)
outlet 284.5'c/7 MPa, with up to 14.5 wt.% steam
return from turbine loop at 155-165'c
separator drum pressure 6.9 MPa, 70 kg/cm2
water must be kept below saturation temperature, to prevent film boiling and associated heat transfer rate drop
too hot inlet water makes pumps prone to cavitation - pressure drop on blades leads to localized boiling
too hot inlet water makes reactor unstable - boils too early, too low, increases reactivity
steam goes to a common manifold, through turbines (or their bypass) to condensators and deaerators, then via feedwater pumps to separators
too little steam from reactor overheats the water in the drums
deaerator tank is main water reservoir
max allowed heat-up rate is 10'c/h, max cooldown 30'c/h - to avoid heat shocks to structure
reactor trip with too high or too low drum level (two levels of alarm), fail of two pumps on the same side, high steam pressure, low feedwater flow; can be disabled
water handling
111 kg/sec of coolant diverted through purification and cooling system
removal of dissolved salts, fission products, corrosive ions
precooled in regenerator to about 68'C
then cooled further to 50'C
then fed to filters
first filter is a cylinder with perlite bed, mechanical filter of particulates and leaked lubricants
boiling degree: voids in water absorb less neutrons, power increases with more boiling
circulation water inlet temperature: how easily it boils with given energy input, where in the reactor it starts to boil, difference to boiling point
water loop pressure: boiling point
control rods: external control of reactor power
control
reactor
control rods
power feedback from ionization chambers in/around core
water
coolant pumps, main loop circulation flow - dialogue with reactor control
feed from deaerators, feedwater return, adjusted by valves
balancing steam production against reactor control, drum levels
balancing hot/cold mix from reactor loop (285'c) vs deaerator (155-165'c) for desired 270'c inlet
turbine
utilizes steam from separators
controls turbine and bypass valves?
not enough data
Instabilities
xenon burn - positive feedback on both power increase and decrease
void coefficient positive - increased boiling increases power
requires enough control rods inserted deep enough (graphite tips) to rapidly decrease reactivity
negative temperature coefficient for fuel - hotter fuel decreases power
however, at low temperatures, low powers below 20% design power, the void coefficient strongly dominates
inlet-outlet temp difference - how gradual boiling is along the zone, at which level
high temperature of water at pump inlet lowers margin to cavitation
control rod design - graphite tip, water displacement, zone of low absorption
positive power coefficient at low power - under 700 MWt increase/decrease of power leads to further increase/decrease and reactor is unstable, above is negative
test
aim (kinetic energy in turbine for generators - test of new exciter control circuit, turbine vibrations)
mobile vibrolab, Kharkov plant
plan
afternoon test
reality
required power, no shutdown permitted
core poisoned
shift change
new shift, night shift not properly instructed - plant has some 4000 parameters to track
turbogenerator coast test
to maintain decay heat cooling, pumps must run
to run pumps, electricity is needed
in case of power cutoff, electricity produced on-site by diesel generators
generators take about a minute to start, worries that the reactor may overheat meanwhile
kinetic energy stored in the spinning turbogenerator, can be extracted to run the pumps - is it enough?
previous tests failed due to premature loss of magnetization current of generator rotor (dynamo on the same rotor) and too fast loss of generator output voltage
at this test, new control circuit added to the magnetization generator
turbogenerator coast test detailed plan
reactor at 700-1000 MWt
TG8 brought online, connected to four recirculation pumps to simulate load
MCP-11,21 on grid standby
MCP-12,22 on grid
MCP-13,14,23,24 on TG8
steam flow to turbine stopped, reactor scrammed
worked before, considered safe; done without incident on unit 3 in 1984, albeit electrically failed
Fomin, chief-engineer, considers it easy; been there done that; doesn't tell State Committee for Nuclear Safety, not even Brukhanov, the plant director
two important changes in the test
all 8 main circulation pumps specified to run
test program sequencer patched into control panel, simulating the design basis accident
reactor safety systems told external power lost; start generators; connect TG8 electrical output to main pumps
test expected to take less than a minute
reactor air cooling test
to be run after shutdown, measure the possibility to dump residual heat to air via surfaces of the separator drums and their water-steam tubes
sink: grid via transformer, virtually unlimited to the power limit of generator
control: excitation generator power to the rotor windings
rpm stabilized by grid
insufficient torque leads to motoring
insufficient load leads to overspeed - critical with grid disconnect
6kV output, can directly feed main circulation pumps - point of the tests
rpm and output voltage critical - pumps trip at 45Hz and below 75% voltage (4.5 kV)
regulated by adjustment of exciter current
timeline
April 25 1986 - scheduled reactor shutdown, tests tacked on
night shift
0106: reactor at 3100 MWt/603'C, OZR 31 rods RR; start of reactor power lowering - power changes are slow to limit thermal shocks to the core
0345: begins switch from helium-nitrogen to nitrogen
0347: power at 1600 MWt
0710: OZR 13.2 RR
morning shift
0800: first day shift, reactor at 50% - 1520 MWt/522'c, unrelated maintenance tasks being done, ECCS being disabled for unknown reason (avoid thermal shock?)
...TG7 powered down, three circulation pumps running from grid; TG8 feeds grid, powers three other pumps
1400: another plant goes offline, Kiev grid controller calls, test postponed; SAOR disonnected from KMPC
1415: (originally planned start of test)
1520: OZR 16.8 RR
afternoon shift
1600: second day shift, power maintained at 1520-1600 MWt/525'c, water at 50,000 m3/h
1850: gear not under test switched over to active transformer T6
2245: reactor at 1600MWt/525'c/50000m3/h
2304: Kiev allows shutdown
2310: powerdown resume starts, OZR at 26 RR (or 23?)
2330: power at 1200MWt
...(m) team of electrical engineers threatens to cancel the contract and return to Donetsk if the test won't start soon
...(m) nuclear safety department physicist not present at all, was told the test was already done
2400: powerdown completed, reactor at 760MWt, turbine 8 gives 200MWe
night shift
midnight: night shift comes in, originally intended to just supervise the cooling of residual reactor heat
unprepared crew, little time to prepare test
desk R, SIUR: Leonid F. Toptunov on the senior reactor control engineering desk, only two months in this position
desk P, senior unit control engineer: Boris Stolyarchuk
desk T, SIUT: Igor Kershenbaum
Aleksandr Akimov, shift foreman, experienced reactor control engineer, supervising the test; knowledgeable but easily pushed around by authority
Dyatlov, in administrative role; deputy chief engineer, pushes forward - if test not done tonight, it will have to wait for another year
good knowledgeable expert but too authoritative, hard to work with; career at installing pressurized water reactors to submarines; exposed to 100 rem in Lab 23 reactor accident
disliked but respected professionally, not respected personally, feared
Dyatlov late to meeting with previous shift, had to yell down Block 3 staff for lack of discipline
Dyatlov yells down Akimov for reading notes of the previous shift and being slow
0005: Dyatlov orders further powerdown, to 200 MWt; possibly thinking lower power level is safer
the regulations did not proscribe operation below 700MWt, but should've (i)
Akimov disagrees, Dyatlov insists he knows better, bystanders hear them arguing, Akimov reluctantly gives order, Toptunov complies (m)
...somewhere here: transfer from local to global mode, operator neglected to tell the system to maintain power, power drops rapidly
0028: power around 500 MWt; switchover from local automatic regulator LAR to global automatic, 1AR, 2AR
... reactor control shifted from local automatic (LAR, individual core regions control, feedback from flux detectors distributed through core)...
... ...to global automatic (1AR, 2AR, feedback from neutron detectors around reactor, as LAR is not reliable at low power)
... Toptunov omits telling the automatic regulator at which power level to run the reactor after switching modes, regulator defaults to last setpoint in the mode - near zero
... ...and reactor obediently goes to zero, within two minutes
... according to other reports, there was no operator omission; Dyatlov later refers to the system not working properly (i)
0028: low water level protection on drum separator turned off (first level)
0028: reactor thermal power drops to near zero (30MWt)
... and Toptunov does nothing for about four minutes, while reactor builds up xenon poisoning
... here the nuclear safety procedures demand reactor shutdown, even if it means aborting the test
... Akimov and Toptunov insert control rods, decide to abort test
... Dyatlov insists he was not present at the moment of power drop
... Toptunov claims Dyatlov enraged and commands removal of more rods from the core to increase power (m)
... or there is no significant discussion (w)
... Toptunov refuses to obey, Dyatlov threatens to find another operator (Yuri Tregub, head of previous shift who watches test, is standing by and watching)
... Toptunov risks career and comfortable life in Pripyat, just when it began... and obeys
... six minutes after power drop Toptunov withdraws rods
... Toptunov has issues with finding right balance of manual rods, pulls predominantly from 3rd and 4th quadrant, Tregub advises which rods to choose to pull
... reactor now highly poisoned
003050: fault signal of 2AR
003135-003246: activation of BRU-K2 of TG-8
003403-003749: emergency level deviation signal of drum separators
003624: pressure setpoint in drum separators from 55 to 50 kg/cm2
0038: reactor power between 0..30 MWt
003932-004335 - DREG program not running, SDIVT (senior computer operator of SKALA) changing magnetic tape for test recording
0042: reactor power at 160MWt
0042: TG8 disconnected from grid, vibrations measured at idle with running exciter
004327: operator disables AZ5 output activated when both TG disconnected
... in accordance with operational and test procedures (i)
004335-004440: signal "1PK up"
004336-005145: emergency level deviation in drum separator
004919-005123: signal "1PK down"
005123: activation of BRU-K1 TG8
005227: emergency level deviation in drum separator
010002: AR2 fault signal
010004: emergency level deviation in left side drum separator
010220: coolant flow at left side increased from 104 to 424 t/h (m3/h) to increase level
0103: reactor power at 200MWt, TG8 disconnected from mains, vibration measurements with disconnected generator
... and the poisoned reactor doesn't want to go higher, almost all control rods are now out, in topmost position; graphite tips all aligned in-plane
... engineers disregard the reactivity margin computer, as it is not always accurate
... 211 rods of 203 are now out of the core
0104: GCN12 (MCP-12, main circulation pump 7) on, as part of test program, but never intended to run all pumps at such low reactor power
010602: emergency level deviation in left side drum separator, water circulation increased from 192 to 1170 t/h to increase level
0107: GCN22 (MCP-22, main circulation pump 8) on, as part of test program
...all 8 pumps now on; 4 fed from grid to supplement the other 4 if they stopped during the test, and to allow running test 2nd time if failed
...after shutown the pumps x1,x2 were supposed to continue cooling the core
...boiling/cavitation margin failing, to as low as 3'c
0109: water circulation lowered to 100 t/h on left and right
010945: emergency level deviation in drum separator
011210-011849: DREG program not running, SDIVT changing tape and fixing parameters for "coast" program
0115: PN-3,4 switched on for recirculation
0118: TG8 synchronized and connected to network (by teletype tape of Skala)
0118: TG8 gives 32.7 MWe
011849: DREG program switched on
011852: signals 1MPA and 3MPA passed, 2MPA issued (by DREG)
011854: AR2 switched off due to malfunction, activated BRU-K1 TG8, protection to increase level in BPG to 2nd limit disabled, protection to increase level in BTS disabled
011910(K): feedwater return increased, now joins recirculation flow just beyond recirculation takeoff at steam drums
...drum level slowly raises, but reactor inlet water temperature drops as as feedwater is much cooler
...feedwater now at 3 times the 200MWt equilibrium rate
...feedwater at inlet too hot, no time to dump heat outside of the loop; now highly susceptible to premature boil
011939: signal "1PK up"
011945(K): boiling stops entirely, reactor operates as PWR for about 2 minutes; no steam voids reduce reactivity, AR control rods withdrawn to compensate
012155(K): reduction of water flow (from 4-times the equilibrium)
012210(K): boiling resumes in core, with dramatic effect on reactivity
012210(K): rod bank AR1 being inserted into reactor, at 20 seconds in 90% length
012225(K): rod bank AR3 being inserted; feedwater at 2/3 of equilibrium for 200 MWt; AR1 responding violently as operator tries to set things up
...: SCRAM circuit for turbine trip switched off
012230: parameters before TG coast written to tape of SKALA; reactivity margin 15RR, IAEA calculated post-mishap to 6-8 deg RR
...reactivity surplus now only 8 rods, water becomes most important neutron absorber in bottom part of the core
...together with too low boiling margin, any small decrease of pressure or increase of temperature causes boiling, absorption drop, localized runaway
...here the shift log ends
012304(K): test begins; oscillograph turned on, emergency steam valve to TG8 closed, reactor would scram here if not bypassed, with rods all in at 012325
...: coolant flow rate slowly starts falling; pressure increase from turbine shutoff in drums lowers voids, decreases power
012310(K): one bank of rods starts exiting reactor due to pressure-increase lowering steam voids and decreasing reactivity
012321(K): rods go back again as pressure pulse effect passes, coolant flow decrease effect takes over, voids increase
012326(K): second rod bank joins going into core
012332(K): third rod bank joins the two
...(m): turbine coasts down to 2300 rpm, time to end test, shut down reactor
012340(b): AZ-5 button pressed (SCRAM)
...(K): SCRAM: AZ5 button pressed, but the rods take about 6 seconds to start having effect
012343(b): signal AZ of reactor control period lower than 20 seconds; signal of high power, power above 530 MWt
012343(K): all three automatic rod banks fully inserted, not sufficient for control
...(K): reactor power now far above designed-for 3200MWt, rise in less than 3 seconds
...(K): one second later reactor reaches above 300,000 MWt
...(m): ...sound like of decelerating turbine, then grows to roar, building starts vibrating; all 8 emergency steam valves open, not enough, blown away
...steam valves are first explosion-like sound heard outside, not bringing much attention as pressure valve opening is common occurrence
012346(b): first pair of GCN pumps off (main circulation pumps)
012346.5(b): second pair of GCN pumps off
012347(b): sudden flow drop (to 40%) of GCN pumps not involved in rundown tests (GCN 11,12,21,22); pumps in tests (13,14,23,24) give unreliable data
012347(b): sudden pressure increase in drums; sudden level increase in drums (water pushed in by steam from reactor); measurement error signals from both autoregulators 1AR, 2AR
012348(b): flow in not involved pumps back in nominal; left on 15% less than before, only 10% of full flow on 24, unreliable data on 23
012348(b): continuing increase of drum pressure (75.2 kg/cm2 on left, 88.2 kg/cm2 on right), increase of level in drums; BRU-K1 and BRU-K2 dump valves activated
...(K): one second later reactor destroyed; KABOOM!
...(m): ...building shakes, oscillations slow but growing in force; sound of moan, then loud bang
012349(b): signal "Pressure increase in RP (reactor vessel); (disturbed(?) PK)", signal "no 48V voltage on SUZ servos", signal "measurement error in 1AR, 2AR"
0124(b): SIUR log: loud bangs, SUZ rods stopped, not reached bottom; SUZ power disconnected"
rods seized at 2-2.5m depth (w) or 4m (m), instead of the entire 7m depth of the core
Toptunov releases electromagnetic clutches, so rods would drop down by their own weight; did not work
according to one of hypotheses, bottom part of the reactor went prompt-critical, shot debris through the refueling tubes and roof up to 2.5-3km altitude
local reactor temperature reaches 4650'c in places, explosion disintegrates core
second explosion caused by steam, or hydrogen, or both; equivalent to about 60 tons of TNT
Pyatachok aka Elena pushed off, tears off tubes, pulls rods, sits sideways across the reactor pit
reactor building torn apart
almost 7 tons of uranium fuel, cladding, graphite powdered into dust and thrown high into atmosphere, now around as hot particles
25-30 tons of fuel and graphite in larger chunks thrown out of the core, igniting on contact with air, starting fires
1300 tons of graphite remaining in the core, burning
Valeri Khodemchuk, pump operator, dead instantly; located near the main circulation pumps, entombed in rubble; pharaohs would envy
Vladimir Shashenok, adjuster from Atomenergonaladka, located under feedwater systems, watching manometer gauges; unconscious, contaminated, scalded; evacuated but died at 6 AM
from now, there are efforts to get water into the reactor core, to prevent it from meltdown
lack of comprehension of the situation even when core guts laying around
Dyatlov assumes emergency CPS protection tank exploded, 110m3 makes considerable damage - reactor is intact, must be cooled
Akimov, Tregub and others sent to open various valves, to run ECCS, to run startup circuit pumps, get water from condensers to the core...
main data source (Karpan) at http://www.physiciansofchernobyl.org.ua/rus/books/Karpan/44.pdf
(K): Herbert Kouts, https://inis.iaea.org/collection/NCLCollectionStore/_Public/19/005/19005803.pdf
fuel disintegrated into fine powder, mixes with coolant, causes pressure surge
pressure pulse drives coolant up and destroys refueling caps, down and destroys piping below reactor, sideways and ruptures at least some coolant channels
second explosion:
steam released into reactor pit, rises 1000-ton lid, shears off pressure tubes, pulls remaining control rods out, lid falls back and rests on side
observers described "fireworks display"
33GW recorded peak power, 300+GW estimated
problems at other plants
Leningrad NPP - broken channel in 1975, localized LOCA, small accident
unit improved
control rods redesigned
high speed scram added
Beloyarsky Unit 2, 1977, half of fuel channels melted, year-long repair
similar accident in 1982 in Chernobyl Unit 1
positive scram effect of the graphite displacers discovered in Ignalina in 1983
chief engineering association informed organizations and all RBMK power plants about intention to impose restrictions on complete withdrawal of control rods from core, but the matter fizzled out (i)
is mitigated by redesigning control rod so when fully retracted there is no water below displacer tip in the active part of the core
changes were planned for Chernobyl as well, but slow correspondence between RBMK designers made it fizzle
https://www.rri.kyoto-u.ac.jp/NSRG/reports/kr79/kr79pdf/Malko1.pdf - The Chernobyl Reactor: Design Features and Reasons for Accident
radiation effects
types of nuclear radiation
alpha
helium nuclei, two positive charges, HEAVY, highly energetic, short-range
stopped by leaf of paper, by layer of dead skin cells
massive but highly localized damage
destructive to living tissues when emitter embedded within
free path in air in centimeters
beta
electrons, one negative charge
light, fairly fast
ordinary radiation-produced ones are of low-ish energy, stopped by thin leaf of aluminium
less pronounced ionizing effect than alpha, longer distance
penetrates skin, goes to low few mm depth, causes sunburn-like injury, beta burns
easy to shield against
higher-energy beta produces Bremsstrahlung, braking radiation - trading beta for gamma, lead shielding may be detrimental
higher-energy beta also produces secondary electrons ("delta rays")
common industry use as accelerators
high-energy beta from accelerators (sterilization, crosslinking...) can pierce through the whole body
gamma
photons, no charge
same as x-rays, higher energy
xray and gamma energies overlap, difference is the mechanism of production - emission by electrons jumping between orbitals (xray) vs emission by nucleus deexcitation (gamma)
highly penetrating, not so ionizing - less energy transferred to material over distance
penetration depth depends on energy - soft (low-energy) is easier stopped than hard (high-energy)
medical xray systems have a strip of aluminium as filter to take out the softest part of the spectrum that wouldn't get through even the soft tissues and only contribute to patient's dose
security/airport xray systems have two detector strips, one bare one with copper-strip shielding, as a crude two-band "color" filter, to better determine object density
whole-body exposure, systemic effects
difficult to shield against, requires thick layers of heavy material; the mass is the key, heavier materials just allow thinner layers
common industry use - xrays, radiography, sterilization, materials (crosslinking), measurements, medicine
neutrons
emitted by nuclear reactions - neutron generators (alpha-n reactions with beryllium, fusors...), nuclear fission - most intense and controllable source
heavy particles, no charge
cause nuclear transmutations - can be caught by the nuclei, converting it per chance to a radioisotope - activation analysis
massive damage to living tissues, about 10 times more than any above
difficult to shield, require light nuclei to slow down and absorb; boron in paraffin is common
massive exposure possible during criticality accidents
positrons
antielectrons, beta+
not commonly encountered, produced by proton-rich radioisotopes (artificial, short-lived)
annihilate quickly with naturally occuring electrons, producing gamma photons of specific energy
not encountered in the wild
protons
medical accelerators, used for cancer radiotherapy
easy to regulate penetration depth, transfer most energy in narrow range of depths - lower radiation injury between surface and target area (cf. "gamma knife")
not encountered in the wild
cosmic radiation
wild cocktail of hard UV, xray/gamma, energetic particles
solar wind, galactic cosmic rays
only at high altitudes and space
varies with solar activity, position; van Allen belts
contamination can originate from naturally occuring nuclides
radon is typical - decays to solid products that stay in lungs and cause local exposure
polonium and lead-110, from cigarette smoke
dose vs dose rate
dose is the amount of absorbed energy
sieverts
dose rate is the speed at which the dose arrives
sieverts per year/day/hour/second
the same dose may have less serious effects if delivered over long time
same as watts (energy per time) vs watt-seconds (joule) or kilowatt-hours (3.6 megajoules - 3600 seconds * 1000 joules)
dose equals dose rate integrated over exposure time
higher than absorbed dose at higher energies, as some of the energy escapes the absorbing volume in the form of bremsstrahlung or fast electrons
equivalent dose
absorbed dose multiplied by RBE (relative biological effectiveness), or W_R, differs by type of radiation
RBE 1 for gamma, beta, muons
RBE 2 for protons, charged pions
RBE 2.5 to ~20 for neutrons, depending on energy, with peak at roughly 0.8 MeV
RBE 20 for alpha, fission fragments, heavy nuclei
for uniform whole-body exposure
"rem", roentgen equivalent man (CGS unit) (older definitions vary), defined in 1976 to be 0.01 Sv (1 roentgen of radiation deposits 0.96 rem in soft tissue)
"sievert", Sv (SI units) - 1 Sv = 100 rem
rem outdated but still in use in the USA
large units, millirem and microsievert are more commonly used
1 mrem = 10 uSv
effective dose
equivalent dose multiplied by W_t, tissue type weighing factor
also rem/sievert
used when irradiation is not uniform
represents stochastic health risk
organs get assigned a number, combination of their sensitivity and body percentage
bone marrow gets disproportionally higher value than their body mass fraction, due to high sensitivity
bone surface gets disproportionally lower value, due to low sensitivity
all weighing factors sum to 1
committed dose
internal effective dose, from ingested/inhaled nuclides
rem/sievert
different coefficients used, otherwise similar to external exposure
calculated from organ or tissue equivalent doses, times Wt, times exposure time (taken to be 50 years for adults, 70 years for children)
absorbed dose measured
...multiplied by RBE (radiation type factor) to get equivalent dose
...and when needed multiplied by the tissue factor
ambient dose equivalent
for monitoring penetrating radiation, dose within 10mm layer of sphere phantom towards source
directional dose equivalent
for monitoring nonpenetrating radiation, dose within 0.07mm layer of sphere phantom towards source
personal dose equivalent
for individual dose monitoring, usually using a dosimeter
health effects
deterministic
high doses, major tissue damage
dose usually measured in rads/grays (not sieverts) - 1 rad = 0.01 Gy, 1 Gy = 100 rad
acute radiation syndrome (ARS)
occurs at high doses (over 0.1 Gy) and dose rates (over 0.1 Gy/hour)
stochastic
low doses, long-term effects, usually cancer
dose measured in rem/sieverts - 1 rem = 0.01 Sv, 1 Sv = 100 rem, 1 mrem = 10 uSv
effect of lowest doses is approximated, using the doubtful linear non-threshold model
radiation hormesis hypothesis, beneficial effects of low-dose radiation to keep the cellular repair systems up and running
at lowest doses difficult to dig the real effects out of the noise
at high natural background areas health effects on population are still not observed
common doses and dose rates
98 nSv "banana equivalent dose", dose from potassium-40 in one standard banana
5-10 uSv dental radiograph
80 uSv one-time average dose within 10 miles from Three Mile Island mishap
400-600 uSv one two-view mammogram
1 mSv limit annual total effective dose for civilians, US
1.5-1.7 mSv annual dose, flight attendants
10-30 mSv full-body CT scan
50 mSv limit annual total effective dose, occupational, US; averages 5.7 uSv/h
68 mSv max one-time dose for Fukushima evacuees
80 mSv half-year on International Space Station
160 mSv yearly dose for smokers of 30 cigarettes per day, mostly from Po-210 and Pb-210
250 mSv half-year trip to Mars, cosmic rays
500 mSv limit annual shallow dose equivalent to skin, occupational, US
1 Sv career limit for NASA astronauts
4-5 Sv LD 50/30 of acute radiation exposure - 50% death within 30 days
4.5-6 Sv lethal doses from Goiania event
5.1 Sv lethal dose for Daghlian, 1945 criticality accident, death 25 days later
10-17 Sv lethal doses in Tokaimura criticality accident; 17 Sv dose survived for 83 days
21 Sv lethal dose for Slotin, 1946 criticality accident, death 9 days later
36-45 Sv lethal dose to Cecil Kelley, 1958 criticality accident, death in 35 hours
54 Sv lethal dose to Boris Korchilov, 1961 reactor cooling failure on K-19 submarine
64 Sv dose over 21 years (350 uSv/h) to Albert Stevens, from 1945 plutonium injection experiment; died in age 79, 1966, from heart failure
yearly doses:
2.4 mSv 0.27 uSv/h yearly global average natural background radiation dose
8 mSv 0.9 uSv/h natural background radiation, Finland
24 mSv 2.7 uSv/h cosmic radiation in aircraft, cruising altitude (~10 km)
130 mSv 15 uSv/h ambient field in most radioactive house in Ramsar, Iran - natural
350 mSv 39.8 uSv/h inside "The Claw", Chernobyl exclusion zone
800 mSv 90 uSv/h monazite sand beach, Guarapari, Brazil, natural
1 mSv/h NRC definition of high radiation area
190 mSv/h peak radiation from Trinity test, 20 miles from ground zero, 3 hours after blast
270 Sv/h typical PWR reactor spent fuel, 10 years old, no shielding
530-650 Sv/h inside primary containment vessel of second BWR reactor, Fukushima
radiation-related accidents
radiation exposure from localized source
mishandling of portable source - iridium radiography
industrial sources or accelerators - failure of interlocks
source stays intact, or is inactive when not powered
most common iridium, cobalt, some medical brachytherapy sources
radioisotope dispersion - diffuse exposure, source spread all around, contamination
reactor accident - Chernobyl, Three Mile Island, SL-1 (minimal, lethal due to mechanical effects of reactor self-disassembly)
material storage accident - Mayak
source tampering - Goiania
"dirty bomb"
nuclear fallout
accidental melting of source in scrap metal
often required extensive decontamination, contamination can be spread around
gift that keeps giving
criticality accidents
rare
energy release self-limiting, all within roughly order of magnitude number of fissions (10^16..10^17)
produced massive short dose of gamma and neutrons
produced small amount of fission products, some (minimal) dispersion possible with liquid sources
neutron activation - nearby metals become radioactive - used in dosimetry
radiation effects on materials
ionization and chemical effects
nonconductive materials become conductive - photoionization - electronics
radiolysis of water - yields peroxide, hydrogen
generation of free radicals - degradation of polymers (depolymerization), lubricants (polymerization)
can be leveraged for radiation polymerization - electron beam curing of inks, impregnation of materials with polymers
frequently used for crosslinking of thermoplastics - polyethylene (PE) to PEX
dislocation of atoms in lattice
neutron induced swelling, Wigner effect
embrittlement, structure coarsening - degradation of metal properties
corrosion effects
thermoluminiscence - dosimetry, archaeology
radiation effects on electronics
photoionization - temporary shorts, latchups (parasitic PNPN thyristor-like structures, resulting current can be destructive, needs powercycle to recover)
degradation of semiconductor structure - loss of gain of bipolars, increase of reverse current in diodes
injection of charge carriers into gate oxides - threshold shift in FETs
change of memory cell status - single event upsets
effect of alpha emitters in packaging and passivation of RAM (DRAM?) chips
effects on sensors - "snow" on image sensors
some effects can partially recover over time - "anneal"
dose-dependent, dose-rate dependent, temperature dependent
higher integration, smaller structures - higher sensitivity to single-event upsets
mitigation techniques exist - "rad-hard" parts, often limited to older-generation chips
concern for aircraft electronics, special concern for satellites, reactor electronics, military where nuclear weapon effects can be encountered - also EMP
failure of robots in Chernobyl
neutrons can activate materials, make the electronics itself a source of its own radiation damage
neutrons can even transmute materials, change composition of semiconductors - exploited for silicon doping with phosphorus for solar cells
radiation effects on living tissues
ionization/chemical effects
production of free radicals
radiolysis of water, generation of peroxide
wild mix of chemical reactions caused by the ions and free radicals - protein crosslinking, DNA damage
can trigger apoptosis - cell dies
massive-enough dieout of cells can be acutely dangerous - gangrenous complications of dead tissue, needs debridement; tissue loss can be hazard - skin loss with beta burns, infections
susceptibility of tissues greatly varies - the more active, the more frequently dividing, the worse
gut lining, bone marrow - especially sensitive
bone marrow loss - white blood cells aberration - immune function loss - especially hazardous with gut barrier and skin barrier loss, infections can kill
used in radiation therapy of cancers, where faster-growing cancer cells are also more susceptible than healthy tissue
healthy tissue can be spared by varying the path of the beam - also gamma knife
can damage cell's machinery - progressive damage leads to cancer
external exposure
no radioisotope enters the body, only the energetic particles/photons
localized or systemic effects
spatial delivery of radiation can vary - beam, point source, diffuse source, internal contamination
area can vary - fingertips, hand, buttock, face, whole body
depth can vary - skin-only (beta burns), limited depth, all the way through
damage can vary
none perceived
erythema, mild irritation that resolves itself
tissue death, with blister or necrosis later
immediate tissue destruction (rare, requires high energies)
effects often delayed by hours, days, weeks
can be from proximity to a "shining" object
can be from own contaminated clothes, skin surface, hair
internal exposure
ingestion or inhalation
contaminated food
inhaled aerosols or gaseous isotopes
less commonly direct entry to body via a wound
total dose depends on element activity, amount, concentration in target organ, residence time (biological half-life)
fate in organism varies by element
strontium and other "bone seekers" have chemistry similar to calcium, accumulate in bones, long biological half-life
caesium is similar to potassium, accumulates in muscle tissue, short biological half-life
radioiodine seeks thyroid - iodide tablets to dilute
other elements can concentrate in different organs
insoluble aerosols accumulate in lungs, soluble ones are then resorbed and move to systemic effects
also depends on the compound - same as "ordinary" toxicology of heavy metals, the atoms are the same, except the tendency to fall apart
some elements are native to body, the chemistry change on decay can wreak havoc on the molecule; eg. phosphorus-32 in DNA backbone converts to stable sulfur and breaks the molecule
acute radiation poisoning
four stages
the higher dose, the shorter the stages
prodromal - nausea, vomiting, anorexia, sometimes diarrhea
the faster the start of nausea, the more severe the rad poisoning
at low dose, prodromal stage may not occur for many hours
at high dose, vomiting can occur within minutes (Slotin)
at very high dose, cognitive impairments, ataxia
latent - symptoms disappear, person feels healthy
usually a few days
manifest - symptoms return, organ damage, immunosuppression
can last for weeks to months
final - recovery or death
death in couple hours at highest doses
recovery in weeks to up to two years
less than 0.5 Sv - no symptoms
0.5-1 Sv - temporary drop in white blood cells
1-2 Sv - prodromal stage hours later, then recovery without disability
2-4.5 Sv - vomiting, diarrhea, hair loss; 2 Gy lethal for 5% in 60 days
4.5 Sv - lethal for 50% in 60 days
6 Sv - lethal for most
ARS stages:
I - 0.5-1.5 Sv - no or minimal symptoms for 48 hours, spontaneous recovery
II - 1.5-4 Sv - haematopoietic syndrome - whole body exposure
bone marrow suppression with drop in white blood cells and platelets
infection and bleeding problems
2.5-4 Sv is LD50
III - 1.5-4 Sv - severe hematopoietic syndrome
bone marrow transplant needed
lower LD50 if resources scarce
IV - 1.5-4 Sv - gastrointestinal syndrome
GI epitel death
severe diarrhea, electrolyte losses
V - over 15 Sv - CNS syndrome
confusion, ataxia, sensory deficits
death within 48 hours
early appearance of CNS disturbances is ominous (Kelley)
absolute lymphocyte count: normal at 2500 cells/mL
measure every 4-6 hours for initial 48 hours
1200 - probably not lethal
300-1200 - significant drop, hospitalize
under 300 - critical
external decontamination
chelation therapy for internal contamination
intervention by nuclide
chemical by nature
tritium: dilute (force fluids through the body)
iodine: KI (force dilution) - useful only at beginning of exposure, balance against side effects
caesium: Prussian blue (sorbent/sequestering agent)
4 June 1945, Los Alamos, nuke pit - subcritical mass of enriched uranium with water reflector
water leaked into polyethylene box with the metal, started working as both reflector and moderator, increasing efficiency
three people with nonfatal doses
21 Aug 1945, Los Alamos, nuke pit - Harry Daghlian
first criticality victim
31 Oct 1956, Idaho National Lab, nuclear jet propulsion reactor HTRE-3 suffers excursion and meltdown due to misconfigured power sensors
16 Jun 1958, Oak Ridge, TN, Y-12, first recorded uranium processing criticality
during leak test fissile solution collects in a 55gal drum, results in 20-minute long event, 8 exposed
5 hospitalized for 44 days
all 8 returned to work
15 Oct 1958, Vinca Nuclear Institute, Yugoslavia - criticality mishap in a heavy-water reactor, one dead, 5 injured; first bone marrow transplants in Europe
30 Dec 1958, Los Alamos
layers form in mixing tank, stirrer powered up, creates vortex, organic layer changes shape, fwoosh
Cecil Kelley gets 36-46 Sv, dies 35 hours later
3 Jan 1961, Idaho Falls - SL-1 mishap
24 Jul 1964, Wood River Junction
uranium solution added to tank with dissolved fuel instead of trichloroethylene extractant
criticality happens, operator gets 100 Gy, dies 49 hours later
second criticality happens 90 minutes later when other operator goes to switch off stirring, two get 1 Gy doses without effects
10 Dec 1968, Mayak
experiments with different solvents for extraction
solvents carried over to a tank not designed for them
improvised attempts to decant the plutonium solution, poured to a wrong-shaped vessel
flash of light and heat, operator drops bottle, runs out
shift supervisor reenters building, attempts to dispose of the solution, triggers large excursion, gets fatal dose
23 Sep 1983, Buenos Aires, research reactor operator gets fatal dose of 37 Gy during rod pattern reconfig with moderating water in reactor
17 Jun 1997, Sarov, Russia, 4850 rem fatality
30 Sep 1999, Tokaimura, Japan, fuel reprocessing
uranyl nitrate solution put into tank not designed for it
shape leads to criticality
two dead, one injured
tickling the dragon's tail
Daghlian
21 Aug 1945, Los Alamos, nuke pit - Harry Daghlian
tungsten carbide bricks added as neutron reflectors around subcritical plutonium sphere
Christy Pit, identical to The Gadget pit
6.2kg of plutonium-gallium alloy
nicknamed "Rufus"
pit almost critical, "-5 cents of reactivity"
two days after planned cancelled third bombing run
4.4kg per brick, manually stacked around the pit
54x54x108 mm (1x1x2 element, element being 2.1/4")
Daghlian worked alone (breach of safety protocols)
security guard at a desk 3-4 meters away, in room
last brick moved into position, held above assembly; neutron counter suggested being close to criticality
brick dropped into the middle of the assembly
about 10^16 fissions
brick could not be pushed away, had to be grasped and moved
stack was partially disassembled
2 Gy (200 rad) in neutrons
1.1 Gy (110 rad) in gamma
total 5.1 Sv
tingling sensation in hand during incident
hand badly burned, high whole-body dose
died 25 days later, of acute radiation syndrome, haematopoietic focus
security guard located 3-4 meters away, within room
0.08 Gy (8 rad) in neutrons
0.001 Gy (0.1 rad) in gamma
felt heat wave, saw blue glow
died 33 years later, aged 62, of acute myeloid leukemia
pit "hot" with fission products
pit ready to be shipped to Tinian, still a bit too hot but not too hot to handle
21 Aug morning, Daghlian builds a box on 7x7-element base
becomes critical at 5 layers, with two bricks on centers of 6th layer
afternoon, builds a box on 6x6-element base
becomes critical at 5 layers
next test planned with 5x5-element base
planned for next morning
breach of protocol to work alone and after hours on potential hazard
after dinner, Daghlian attends lecture at theatre #2
lecture ends at 2110
Daghlian arrives to Omega Site at 2130
builds the base, cube
four layers built quickly
fifth layer getting close to criticality, neutron flux increasing
half of the fifth layer finished
holding brick in left hand, moving it near center of cube, counters increased ticking
brick quickly jerked upwards, slipped out of fingers, falls onto the cube
assembly gets supercritical
grabs brick, drops it again
attempts to overturn table, scatter pieces, break geometry - table too heavy
removes bricks one by one until criticality stopped
time is 2155, t0+1 minute
Daghlian stares at the assembly for a while, in shock, assessing the situation; then partly disassembles the stack to render it safe, briefs security guard
t0, tingling felt in right hand immediately as it is exposed to neutron flux
Daghlian driven to Los Alamos hospital
on arrival hand swollen and numb
t0+1.5 hours, extreme nausea/vomiting begins, persists for two days
t0+36h, circulatory system in hand fails, arm turns blue, skin of face and torso turns red
t0+10d, nausea returns, unable to eat; high doses of penicillin, blood transfusions, vitamin B1, quinidine sulfate
t0+25d, coma, later death
right hand exposure estimated 200-400 Gy (20,000-40,000 rem)
left hand exposure estimated 50-150 Gy (5,000-15,000 rem)
body exposure estimated 5.9 Gy (5,900 rem)
security guard in hospital too
feeling slightly tired, temporary increase of leucocytes
released at t0+2d
t0+33y, dies of acute myeloid leukemia
Slotin
21 May 1946, Los Alamos
same "Rufus" pit as in Daghlian's case
hollow beryllium half-spheres used as neutron mirror (then called "tamper")
outer diameter 9", 229mm
bottom half-sphere larger, dia 330mm
top beryllium half-sphere has a thumbhole on its top, to facilitate holding it
distance is slowly decreased, neutron detectors are monitored
if top halfsphere closes with bottom one, assembly becomes supercritical
experiment protocol requires shims to prevent this
Slotin's variant of experiment does not use them
Slotin's variant uses a flat screwdriver to maintain the half-spheres distance
allows getting to criticality faster and easier
Slotin demoes the test for about a dozen times
Fermi mentions he'd be dead in a year if he continues doing so this way
Slotin, ready to leave the lab to teach, teaches his replacement, Alvin Graves
seven people watching test
their dosimetry badges locked in a lead box 30m away
sphere held in left hand, thumb in thumbhole, screwdriver in right hand
1520, screwdriver slips a fraction of inch to the outside
spheres get too close
assembly gets critical - blue glow, heat wave
Slotin jerks left hand up, drops sphere to the floor
t0, sour taste in mouth, intense burning sensation in left hand
t0+1 second, the main event is over
"well, that does it"
Slotin orders others to not move, draws outlines of their feet on the floor
t0+few minutes, vomiting
rushed to hospital
"three-dimensional sunburn"
t+1d, hands swollen and blistered and unusable
latent stage of about two days
t+5d, purple skin, organs failing, high fever, pain
t+8d, coma
t+9d, death
Slotin's dose 10 Gy (1000 rad) in neutrons, 1.14 Gy (114 rad) in gamma - about 21-22 Sv
Graves's dose 1.66 Gy (166 rad) in neutrons, 0.26 Gy (26 rad) in gamma - about 3.6 Sv, 50% chance of survival
Graves partly shielded by Slotin
acute radiation poisoning, loss of hair, zero sperm count
two weeks in hospital, several weeks of convalescence, full recovery
in few months skiing again
bald patch, child with his wife two years later
chronic neurological and vision problems (cataracts)
t0+19y, dies at age 53 of myocardial infarction while skiing, likely combo of hereditary and radiation damage
Kline's dose was 2.5 Gy (250 rem)
nausea and vomiting at t0+few hours
weak, anorexic, nauseated for two weeks
blood pressure drops, immunosuppression, bouts of fever
weak and exhausted for months
skin sensitive to sun
t0+55 years, death
others, leaving hosptial within a week
Young, 0.51+0.11 Gy (51+11 rad), 1.13 Sv, death t0+29y of aplastic anemia and bacterial endocarditis
Cleary, 0.33+0.09 Gy (33+9 rad), 0.77 Sv, death t0+4y by KIA in Korean War
Cieslicki, 0.12+0.04 Gy (12+4 rad), 0.28 Sv, death t0+19y of acute myelocytic leukemia
Schreiber, 0.09+0.03 Gy (9+3 rad), 0.21 Sv, death t0+52y of natural causes
Perlman, 0.07+0.02 Gy (7+2 rad), 0.16 Sv, death probably t0+42y
"Rufus" nicknamed "Demon core"
criticality experiments moved to remote control
"The danger is not that this [cancer] will happen to you. The danger is that it is more likely to happen to you. Maybe the more likely is not very much more likely, but it is still more likely." -- Graves
Kelley
30 Dec 1958, Los Alamos
plutonium recovery by organic-water extraction
various waste gets dissolved, extracted, purified
tank supposed to contain "lean" solution (less than 0.1g Pu per liter) in nitric acid
due to two improper transfers into the tank before, there was nearly 20g Pu per liter, 200 times more
plutonium concentrated in organic phase, floating on water phase, almost 3kg Pu total, already nearly critical
Kelley looking into the tank through port window
stirrer switched on, water forms vortex, organic phase forms a plano-convex "lens", with aqueous reflector/moderator below
about a second later, criticality is reached; t0, lasts for about 200 milliseconds
brief flash of blue light, thud
within three seconds the layers are intermixed, criticality condition is self-mitigated
Kelley collapses or is knocked off the ladder, recovers, switches stirrer off and on again, runs out of building
found outside in snow, in state of ataxia, repeating "I am burning"
other workers switch off stirrer, take Kelley to decon shower
assumption is on mishap of chemical nature, excursion in mixing tank is not considered
up to t0+100 minutes, Kelley is incoherent, waves of vomiting, then period of stabilization
blood assay shows 9 Gy from fast neutrons, 27 Gy from gamma
t0+6h, total loss of lymphocytes
t0+24h, bone marrow biopsy shows watery marrow without red blood cells