Diagnostic Shielding Methodology: NCRP Report 147

Diagnostic Shielding Methodology: NCRP Report 147

Structural shielding design in diagnostic radiology ensures that staff and members of the public are protected from radiation produced during medical imaging. In 2004 the National Council on Radiation Protection and Measurements published NCRP Report No. 147 (Structural Shielding Design for Medical X-Ray Imaging Facilities), superseding the x-ray imaging recommendations of NCRP Report No. 49 (1976). NCRP Report 147 replaced single-kVp assumptions with measured clinical workload distributions, moved from exposure to air kerma ($K_a$, in grays), and introduced broad-beam transmission fits that account for spectral hardening and scatter buildup in real barrier materials.

---

Shielding Design Goals ($P$)

The shielding design goal ($P$) is the air kerma allowed at the occupied point beyond a barrier, stated per week because workloads are traditionally weekly:

NCRP Report 116 recommends an occupational limit of $50\text{ mSv}$ per year with a cumulative limit of $10\text{ mSv}$ times age, and advises that new facilities be designed for a fraction of the implied $10\text{ mSv/yr}$. Report 147 chose half of that value, $5\text{ mSv/yr}$, expressed as $5\text{ mGy}$ air kerma ($0.1\text{ mGy/week}$), which also lets pregnant radiation workers continue in their usual work areas. Because the method ignores patient attenuation, assumes perpendicular incidence, and uses conservative occupancy, the actual effective doses behind a barrier built to $P$ are normally well below these values.

---

Shielding Parameters: Workload, Use, Occupancy, Distance, and Preshielding

1. Workload ($W$) and Workload Distribution

Workload is the time integral of tube current, in $\text{mA}\cdot\text{min}$ per week. NCRP 147 also specifies how that workload is spread across tube potentials (the workload distribution) using the AAPM Task Group 9 clinical survey, because a $70\text{ kVp}$ exposure needs far less shielding than a $120\text{ kVp}$ one. Workload is usually entered as patients per week ($N$) multiplied by a normalized workload per patient. Table 4.5 of the report gives, at $1\text{ m}$, the unshielded primary air kerma per patient ($K^1_P$) for each distribution:

2. Use Factor ($U$)

The use factor is the fraction of the primary-beam workload directed at a barrier. NCRP 147 Table 4.4 (from the TG-9 survey) gives 0.89 for the floor, 0.09 for the cross-table wall, 0.02 for any other wall, and 1.00 for the chest image receptor wall (using the chest-bucky workload). The ceiling and control booth are generally treated as secondary barriers. For any secondary barrier, $U = 1$, because scatter and leakage reach every barrier during every exposure. Because fluoroscopic image receptors and the mammography breast support are required to act as primary beam stops, those rooms use $U = 0$ for primary radiation.

3. Occupancy Factor ($T$)

The occupancy factor is the fraction of the beam-on time that the single most exposed person spends in the area, not the fraction of time anyone is there. NCRP 147 Table 4.1 suggests:

These are guidance values; the physicist should use real occupancy when it is known and remember that a storage room can later become an office.

4. Distance ($d$)

Distance runs from the source to the nearest likely position of a person's sensitive organs: at least $0.3\text{ m}$ beyond a wall, $1.7\text{ m}$ above the floor of an occupied room below the x-ray room, and $0.5\text{ m}$ above the floor of an occupied room above it.

5. Preshielding ($x_{\text{pre}}$)

Primary calculations ignore attenuation by the patient but may credit the hardware the beam must cross first. NCRP 147 Table 4.6 gives equivalent thicknesses of $0.85\text{ mm}$ lead ($72\text{ mm}$ concrete, $7\text{ mm}$ steel) for the grid, cassette, and support structures of a table or wall-mounted image receptor, and $0.3\text{ mm}$ lead ($30\text{ mm}$ concrete, $2\text{ mm}$ steel) for cross-table work where only the grid and cassette intercept the beam. The credit applies only where the receptor actually shadows the barrier.

---

Primary & Secondary Barrier Formulations

Primary Barriers

For $N$ patients per week, the unshielded weekly primary air kerma at the occupied point is:

$K_P(0) = \frac{K^1_P \, U \, N}{d_P^2}$

and the transmission the barrier must provide is:

$B_P = \frac{P / T}{K_P(0)} = \frac{P \, d_P^2}{K^1_P \, U \, N \, T}$

The total thickness found from $B_P$ includes any preshielding, so the barrier itself needs $x_{\text{barrier}} = x_{\text{total}} - x_{\text{pre}}$.

Secondary Barriers

Secondary radiation is patient scatter plus tube-housing leakage. NCRP 147 Table 4.7 tabulates the unshielded secondary air kerma per patient at $1\text{ m}$ ($K^1_{\text{sec}}$) for each workload distribution; leakage is modeled at the regulatory limit of $0.876\text{ mGy}$ ($100\text{ mR}$) per hour at $1\text{ m}$ (21 CFR 1020.30(k)). The weekly secondary kerma is $K_{\text{sec}}(0) = K^1_{\text{sec}} N / d_{\text{sec}}^2$, and the required transmission $B_{\text{sec}} = (P/T)/K_{\text{sec}}(0)$ is converted to thickness with combined secondary-radiation transmission fits (Appendix C of the report). This replaced the NCRP 49 practice of computing scatter and leakage thicknesses separately and adding one half-value layer when they were similar.

---

Broad-Beam Transmission, the Archer Model & Materials

Narrow-beam attenuation ($I = I_0 e^{-\mu x}$) does not describe a real barrier: scattered photons build up inside the material, and the polyenergetic beam hardens as soft photons are absorbed in the first layers. NCRP 147 therefore fits broad-beam transmission with the Archer equation:

$B = \left[\left(1 + \frac{\beta}{\alpha}\right) e^{\alpha \gamma x} - \frac{\beta}{\alpha}\right]^{-\frac{1}{\gamma}}$

which can be inverted for the thickness:

$x = \frac{1}{\alpha \gamma} \ln\left[\frac{B^{-\gamma} + \frac{\beta}{\alpha}}{1 + \frac{\beta}{\alpha}}\right]$

The parameters have a clear meaning. For large $x$, $B \approx (1 + \beta/\alpha)^{-1/\gamma} e^{-\alpha x}$, so $\alpha$ is the attenuation coefficient of the hardened beam deep in the barrier, and $\ln 2/\alpha$ is the half-value layer at high attenuation. At $x = 0$ the slope is $\alpha + \beta$, so $\alpha + \beta$ describes the faster initial attenuation of the soft beam; $\gamma$ is a fitted shape parameter.

The same fits give useful lead equivalents for a $100\text{ kVp}$ primary beam: one $5/8$-inch layer of gypsum wallboard is worth about $0.05\text{ mm}$ of lead, two layers about $0.10\text{ mm}$, $25\text{ mm}$ of concrete about $0.27\text{ mm}$, and $100\text{ mm}$ of concrete about $1.5\text{ mm}$.

Sheet lead is sold by weight per square foot; NCRP 147 notes that the weight in $\text{lb/ft}^2$ equals the nominal thickness in inches multiplied by 64:

---

Worked Calculation: Chest-Bucky Primary Barrier

A radiographic room performs $N = 120$ chest-bucky examinations per week. The wall behind the chest bucky borders a corridor (uncontrolled, $P = 0.02\text{ mGy/week}$, $T = 1/5$). The distance from the focal spot to $0.3\text{ m}$ beyond the wall is $d_P = 2.2\text{ m}$. For the Rad Room (chest bucky) distribution, $K^1_P = 2.3\text{ mGy}$ per patient and $U = 1$; NCRP 147 Table B.1 gives lead fitting parameters $\alpha = 2.264\text{ mm}^{-1}$, $\beta = 13.08\text{ mm}^{-1}$, $\gamma = 0.5600$.

1. Unshielded weekly primary air kerma:

$K_P(0) = \frac{2.3 \times 1 \times 120}{(2.2)^2} = \frac{276}{4.84} = 57.0\text{ mGy/week}$

1. Required transmission:

$B_P = \frac{0.02 / 0.2}{57.0} = 1.75 \times 10^{-3}$

1. Invert the Archer equation ($\beta/\alpha = 5.777$, $\alpha\gamma = 1.268\text{ mm}^{-1}$, $B^{-\gamma} = 34.95$):

$x_{\text{total}} = \frac{1}{1.268} \ln\left[\frac{34.95 + 5.777}{6.777}\right] = \frac{\ln(6.009)}{1.268} = 1.41\text{ mm Pb}$

1. Subtract the $0.85\text{ mm}$ preshielding provided by the bucky: $x_{\text{barrier}} = 1.41 - 0.85 = 0.56\text{ mm}$, so $2\text{ lb/ft}^2$ lead ($0.79\text{ mm}$) meets the goal behind the receptor. If the same wall section also receives cross-table exposures or the receptor does not shadow it, no preshielding credit applies and $1.41\text{ mm}$ requires $4\text{ lb/ft}^2$ ($1.59\text{ mm}$).

---

Clinical Scenarios & Health Physics Pitfalls

Clinical Transition Scenario

A clinic converts an ultrasound room into a radiographic suite, and the architect submits plans based on NCRP Report 49 (uncontrolled limit $10\text{ mR/week}$, about $0.087\text{ mGy}$ air kerma, and a corridor occupancy of $1/4$). That gives $P/T \approx 0.35\text{ mGy/week}$ at the corridor. Under NCRP 147 the same corridor needs $P/T = 0.02/(1/5) = 0.1\text{ mGy/week}$, so the legacy design would allow about 3.5 times more transmission. The qualified expert should redo the design to NCRP 147, which most state shielding-plan reviews now expect.

Critical Health Physics Pitfalls

• Applying use factors to secondary barriers: Scatter and leakage reach every wall during every exposure, so $U = 1$ for secondary barriers.
• Using the first-layer HVL for every halving: The soft beam is attenuated quickly at first and more slowly deeper in the barrier; thickness must come from the broad-beam fits, not from repeated use of an initial HVL.
• Measuring distance to the wall face: Distance is taken to $0.3\text{ m}$ beyond the barrier (or $1.7\text{ m}$ above the floor below and $0.5\text{ m}$ above the floor above); measuring to the inner wall surface overstates the inverse-square benefit.
• Taking preshielding credit where the receptor does not shadow the wall: The $0.85\text{ mm}$ credit applies only to the area actually behind the image receptor.