Michael H. Fox
Why We Need Nuclear Power
The Environmental Case
Read Chapter 7 for more information.

A dose of radiation is the amount of energy that is absorbed from the
radiation by matter.  The most basic definition of dose is
absorbed dose
which is given in units of gray (Gy) or an older unit of rad (radiation
absorbed dose).  One Gy of radiation is one joule of energy absorbed in
one kilogram (kg) of matter.  It makes no difference what kind of matter for
this definition.  One Gy is 100 times as large as one rad.  An absorbed
dose D is independent of the type of radiation; it is just an amount of
energy absorbed per unit mass.  

Since we are usually more concerned about the effects of radiation on
biological systems, e.g. cells or DNA, absorbed dose is not an adequate
measure of dose.  This is because various forms of radiation can cause
different amounts of damage in biological systems for the same
absorbed energy.  For example, an alpha particle or a neutron is much
more damaging to DNA than an electron (beta) or gamma (γ) or X-ray.  
This is because it has a higher density of ionization as it moves through
A more appropriate definition is the equivalent dose (H) which is given in
sieverts (Sv).  Each type of radiation is given a radiation weighting factor (Wr)
which is multiplied by the dose D to give the equivalent dose H.  For
example, an alpha particle has a Wr of 20, so it is 20 times more damaging
to DNA than a gamma.  Thus, an absorbed dose of 1 Gy of alpha particles is
equal to an equivalent dose of 20 Sv, while 1 Gy of gammas or electrons is
equal to 1 Sv.  But 1 Sv of alpha particles will cause the same amount of
damage to DNA as 1 Sv of electrons or gammas.
Absorbed Dose
gray (Gy)
1 joule/kg
100 rad
Absorbed Dose
1 erg/gm
0.01 Gy
Equivalent Dose
sievert (Sv)
millisievert (mSv)
0.001 Sv
100 rem
0.1 rem
Equivalent Dose
millirem (mrem)
0.001 rem
0.01 Sv
10 μSv
Table 7.1.  Various methods to specify dose
Table 7.2.  Radiation weighting factors (ICRP*)
That is not the end of it either, because different tissues have different
sensitivities to radiation.  Another factor is introduced -- the tissue weighting
factor, Wt -- to take this into account.  This is the basis for the
effective dose
It has the same symbol as equivalent dose, but is equal to equivalent
dose H times Wt.  For example, lungs are about 12 times more sensitive to
radiation than skin, so the effective dose to lungs of 1 Sv equivalent dose
would be 12 times higher.   The effective dose to a whole person is the sum
of the individual effective doses to the organs.
Type of Radiation
Radiation Weighting Factor (Wr)
Photons (X and γ)
Electrons (β)
α-particles, fission fragments, heavy
depending on energy
*ICRP:  International Commission on Radiological Protection is an
international organization that makes recommendations on radiation
protection standards.
Total Wt
Bone marrow, breast, colon, lung, stomach
Bladder, esophagus, gonads, liver, thyroid
Bone surface, brain, kidneys, salivary glands, skin
Remaining tissues
Table 7.3.  Tissue weighting factors (ICRP*)
Because radiation ionizes atoms and molecules, it can cause damage to
cells.  In particular, it can cause both single-strand and double-strand
breaks in DNA (see Figure 7.1 above).  The majority of this damage gets
repaired by DNA repair systems in cells.  DNA double-strand breaks are the
type of damage that is most likely to kill cells or cause changes that may
ultimately lead to cancer.  There is a long latency period of many years
before a cancer can arise from a dose of radiation.  For solid tumors (except
thyroid tumors), the latency period is from about 10 years to 60 years; for
leukemia it is about 5 years but the risk is over by 15 years.  This long
latency period is due to the multistep nature of cancer.  Radiation is an
initiating event, but it takes many more genetic changes over a period of time
to convert a normal cell into a cancerous cell.  

Radiation biologists and epidemiologists have a good understanding of the
relationship between a dose of radiation and the chance it will cause
cancer.  This is based on studies of cells and animals, but ultimately
depends on studied of human populations that have been irradiated.  Some
of this is from people who have been treated or diagnosed with X-rays or
people who have been exposed to nuclear accidents.  But the most critical
data comes from the Japanese people who survived the atomic bombs
dropped on Hiroshima and Nagasaki in 1945.  Over 120,000 survivors in the
two cities have been studied ever since in the Life Span Study.  Individual
dose estimates are known for 93% of the people in the study.  The incidence
of cancer and all causes of death are followed rigorously for all of the people
in the Life Span Study.  Importantly, a control population of over 25,000
people who were not in the cities at the time of the bombing is used to
determine the natural rate of cancer in a similar population.  For example, as
of 2003 (the latest year of compilation), out of 48,000 people who had a dose
of 5 mSv or more, 6,308 had cancer but only 525 of those could be attributed
to the radiation.  

Based on scientific assessments of the Life Span Study and other studies,
the ICRP concludes that the risk of dying of cancer from radiation for
doses below 200 mSv is 4% per Sv for adults and 5% per Sv for a
population that includes children, who are more sensitive to radiation.  For
larger doses, the risk is doubled to 8% and 10% per Sv respectively.

The reason the risk is lower for a lower dose is because cell and animal
studies show radiation is less effective in causing cancer when is given in
smaller doses.  At very low doses of a few mSv, it is not possible to actually
measure an effect, so models are used to extrapolate the results from
higher doses.  The linear, no-threshold (LNT) model is the accepted model
for extrapolating from high to low doses.  However, this is an area of active
scientific study and debate.  There is evidence that this overstates the case
because of the phenomenon of
hormesis or understates the case because
of the phenomenon known as the
bystander effect.  Both of these
phenomena can be demonstrated in cellular studies but the scientific
consensus is that the LNT model is the conservative model to use so as not
to expose a population to an undue risk.  
Figure 7.1.  Representation of energy deposited in DNA from
an X-ray or gamma compared to an alpha particle.