Register for Updates | Search | Contacts | Site Map | Member Login

ICRP: Free the Annals!

View Comment

Submitted by Albrecht Kellerer, ICRP C1
   Commenting as an individual
Document 2005 ICRP Recommendation
 
PART 3 OF 3
(S16): Eq(S1) is reasonably close to the more complicated equation given in ICRP 92 (foot-note on p.78). However, it may be difficult to identify with adequate precision the corresponding function w(L). Also, both Eq(S1) and the equation in ICRP 92 are realistic only up to neutron ener-gies below 100MeV. For the higher neutron energies at aviation altitude, in space, or in the vicinity of high energy accelerators – situations where radiation protection becomes increasingly important – computations support a radiation weighting factor substantially less than 5 (see also paragr.(211)). This can be accounted for by changing Eq(S1) to:
wR = 2.6 + 16 exp[ln(En))2 /4] + 4 exp[ln(En/20))2 /12] (S1)
(in the subsequent line write: “where En is the neutron energy in MeV”)
The specific form of the equation is less important than the fact that all wR values need to be consis-tent in the sense that they correspond to the same specified LET-dependent weighting factor, w(L). (see also comments to (70) and proposed modification).
The mean wR value for a slightly degraded fission-neutron spectrum, changes little against the cur-rent values; but for a highly degraded spectrum it is substantially decreased.

Fig. S1 then takes the form:
Figure 1. Radiation weighting factor, wR, for incident neutrons versus neutron energy.
(A) Step function given in Publication 60, (B) continuous function proposed in this report


(the continuous approximation offered in Publication 60 can be, but need not be inserted)


(S16), 3rd last line: “ where En is the energy of the incident neutrons in MeV.”

(Some of the above comments are taken up again with regard to Section 3.)

Comments on Chapter 3

(33) last sentence: How can the operational quantities ambient dose equivalent and personal dose equivalent be used for “ investigating situations involving ..... intakes of radionuclides ” ?

(35): Should not “ ´amount´ of radiation” be “ ´magnitude´ of exposure” ?
Or, at least, replace “radiation” by “irradiation”.

(37): This defense of the “gross (macroscopic) quantities” does not hurt, but it may not actually be required. As stated at the end of (47), the “discontinuous nature of the physical and biological processes of ionisation“ is accounted for through the radiation weighting factor wR or the LET-dependent weighting factor Q(L). Why should there then be a need to use “stochastic” quantities to specify doses to tissues or organs?

(43) d is termed “mean energy imparted .. in a volume element ” while in (44) it is termed “expectation value of the stochastic quantity  “. In the absence of more detailed explanations it may be better to use the latter formulation both times.

(45) and (46) are useful, because they explain important characteristics of low and high LET radiation, but (44) and (47) may be expendable. These sections are somewhat problematical, because they deal all too briefly with intricate issues. For example: As with other quantities, it depends on the context whether LET is a “stochastic” or a “non stochastic” quantity. Usually when a cell is traversed by certain charged particles the LET is a stochastic variable, it varies ran-domly as does the energy of the particle. The exception are track segment experiments where the LET has a fixed value. The term “non stochastic” in (47) refers apparently to the difference between L and the microdosimetric variables y or . But this issue would need to be explained in more detail, if it were actually required in the present context.

(44): 2nd and 3rd sentence:
Is there a meaningful distinction between a “fundamental quantity” that has “the scientific rigour” and a “fundamental” quantity that does not? All quantities need to be clearly defined, and any clearly defined quantity can, in principle, be determined by measurement, computation, or a combi-nation thereof. I do not wish to dwell on semi-philosophical issues, but the assertion that effective dose is not “measurable” has created considerable confusion. A measurement may be too difficult to be practicable, but new techniques may change this, unless its perceived impossibility is turned into an axiom. As Thomas recommends in his Commentary on ICRP 92 (Aug.10, 2004 draft sent to the Chairman of ICRP, p.4), it might be best “to leave it to the ingenuity of dosimetrists to deduce the means of measurement”.

(45): The first sentence is not sufficiently clear. Different meanings might be intended, e.g.:
- The actual value of energy imparted in a cell (the elementary unit of life) equals the number of energy deposition events times the average energy imparted per event.
(Note: This applies independent of absorbed dose. “Frequency” can not stand for “number”)
- At a given absorbed dose, the expected energy imparted in a cell is given by the product of the expected number of energy deposition events times the expected energy imparted per event.
(Note: I doubt that “frequency” can stand for “expected number”)

(46), 2nd line: Replace “frequency” by “number”

(48), first sentence: “averaged in time” is not correct and probably not needed. The intention is apparently to state that absorbed dose relates to a specified time interval.
line 7: perhaps “meaningful” rather than “possible”

(51): see earlier caution (S14) against another new special name for the unit J/kg.

p.18, footnote: This reasoning does not appear to be helpful. Operational quantities ought to be useful approximations to E. To freeze the operational quantities to their current form – and to ignore the gap between them and E – will preserve the artificial distinction between “computable” and “measurable” quantities. Again to quote the Commentary on ICRP 92 by Thomas: “It may not be wise to endorse ambient dose equivalent to the exclusion of alternatives.” The longer a dual system of dosimetric quantities is preserved, the more damaging will be its effect and the more awkward will be the ultimate return to a coherent system of dosimetric quantities.
(53), last sentence:
The numerical adjustments are necessary. But to ignore the central issue, the restoration of a linkage of wR to a local LET-dependent weighting factor is wasting the chance to return to a coherent sys-tem of quantities (see subsequent proposed modification of paragraphs (70-72)).
(55):
The paragraph is ok except for the first sentence: “Effective dose is ...... in principle as well as in practice a non-measurable quantity”. A non-measurable quantity is a paradox. The measurement of E is too complex to be done routinely, but it is possible in principle. The low level of ambition expressed in the first sentence is particularly astounding when contrasted to (35) where it says:
“Ideally, for demonstrating compliance with the constraints, there would be one single dosimetric quantity ...” ... “The Commission has introduced such a single quantity, the effective dose, as an approach to overcome some of these problems. This quantity can be used for regulations of impor-tant parts of health effects.”
(I wonder, why the notion of E being “non-measurable” has been so willingly accepted. Is it because the live human body is still felt to be somewhat outside the realm of physics proper? Or could it be that – in a sub-liminal understanding – the field of dosimetry is being divided up into the “computational” territory of ICRP and the “measurable” territory of ICRU?)

(57), line 7: “precisely” instead of “precise”

(61): Was the 1991 choice of wR based on radiobiological findings that provided different RBE values than those that determined the 1991 choice of Q(L)? Why should different radiobiol-ogy evidence have been used for the two conventions simultaneously introduced in Report 60?
It is clear that wR corresponds, for neutrons, to larger RBE values than does Q(L) (by about a fac-tor 1.6, even after the obvious reductions of wR are made below 1MeV). But whether the difference is intentional or has merely been the result of some numerical approximations, remains unclear. Unless the discrepancy between Q(L) and wR is removed, the issue must be clarified.
(Line 5: What are the “in vivo investigations on cells”? Why not specify the system(s)? )

(65): The argument is suggestive but perhaps insufficiently quantified. A quantitative analysis shows that the degraded field from 1MeV -rays in a very large absorber still has a kerma weighted mean initial electron energy 380keV, while the corresponding value is only 30keV for ortho-voltage x-rays. For the degraded spectrum from the A-bomb -rays the kerma weighted mean electron energy is even considerably larger.

(70 - 74): If Eq.(S1) is changed so that it provides lower values of wR at very high neutron en-ergies, the subsequent formulation can be substituted for (70) to (74).

Proposed text for paragraphs (70) to (72) (to replace (70) – (74)):
(70) The calculation of the energy dependence of the radiation weighting factor can be based on an LET-dependent weighting factor, w(L), that parallels the Q(L) relation-ship defined in Publication 60 (ICRP, 1991a) but reflects the fact that wR corresponds to somewhat larger RBE values. With L in keV/m:
= 1 L<=10
w(L) = 0.5L  4 for: 10 <= L <=100 (3)
= 420/ L 100 <= L
For a standard anthropomorphic phantom (average over male and female) and isotropic inci-dent radiation a mean weighting factor then results that can be used as radiation weighting factor:
wR = T wT w(L) DT(L) dL / T wT DT (4)
(DT(L): distribution of absorbed dose to organ T in LET)
Since the different organs and tissues are not symmetrically distributed in the human body, the organ absorbed doses depend on the directional incidence of the radiation. Eq(4) provides, therefore, somewhat different values when the incident radiation field is not isotropic. How-ever the differences are insignificant, especially at high energies.
(71) The values obtained in terms of Eq.(4) are given very closely by the following nu-merical dependence on the energy, En , (in MeV) of the incident neutrons:
wR = 2.6 + 16 exp[ln(En))2 /4] + 4 exp[ln(En/20))2 /12] (5)
This function is recommended to replace the function which has been introduced in Publication 60. Figure 1 compares the two relations.
Figure S1: Radiation weighting factor, wR, for incident neutrons versus neutron energy. (A) Step function given in Publication 60, B) continuous function (Eq.(4)) recommended in this report.



(72) The definitions in Eqs(1) and (2) for radiation weighted dose, HT, and for ef-fective dose, E, require the subdivision of the incident field into radiation types and en-ergies and the determination of their separate dose contributions. With common radia-tion fields this presents no problem, but with complex high energy fields – near accel-erators, in aviation altitude and in space – it can be difficult. An alternative approach that provides essentially the same values, but does not require spectroscopy of the inci-dent field can then use the LET-dependent weighting factor w(L) specified in Eq.(3):
E = T wT HT with: HT =  w(L) DT(L) dL (6)

High energy radiation fields are often in near equilibrium, which implies that the radiation quality remains roughly constant throughout the human body. The mean value of the LET-dependent weighting factor can then be determined at a point:
=  w(L) D(L) dL / D (7)
which provides the simple relations:
E = T wT HT with: HT = DT (8)


(Adjust paragraph numbers accordingly; paragraph (75) becomes (73) etc. )

Continuation of specific comments:
(75), last 3 sentences and (76), last sentence:
wR is correctly stated to correspond to different RBE-values than Q(L). It would, therefore, be awk-ward to derive wR values directly from Q(L). Instead, there should be reference to Eq(6) in para-graph (72):
“ .... The mean value is calculated to be less than 1.4 for 100 MeV protons stopping in tissue. At very high proton energies, near 1 GeV, secondary charged particles from nuclear reactions become more important and the mean weighting factor increases up to =2.3. Taking all considerations and available data into account, the radiation weighting factor for protons of all energies should have a value of 2 (publication 92; ICRP, 2003c).”

“ ....From calculations using the w(L) function a radiation weighting factor of about 20 is es-timated.”

(84), 2 last sentences:
It would be better to write in the second last sentence “ ... and generally is not measured.”
and in the last sentence: “... to measured physical quantities, ..”

(86): The text does not quite say what it intends to convey and therefore needs to be reformulated. It is not true that E is defined for a hypothetical reference individual. This would only be so, if E was expressly defined for a standard phantom. The effective dose to John Dow is the E to this particular person. The fact that wR and wT are standard parameters not tailored to J.D. merely shows that – very reasonably – the quantity E is not de-signed to account for all subtle or unknown details of the person. The organ absorbed doses vary much more critically from person to person than do the weighting factors.

(90): Remove the bar over DT. There is no bar over DT in all earlier equations. Organ dose DT and radiation weighted organ dose HT are understood to be averages.

(211), 6th line:
This is quite correct. But note that “neutrons” applies only, if – in line with the present pro-posal – Eq(8) and Fig.1 are changed to lower values of wR at very high neutron energies.