*Core Reload Analysis Techniques in the Advanced Test Reactor DOI: http://dx.doi.org/10.5772/intechopen.103896*

peak-to-average factor for fuel element five (shown in **Figure 1**) in the northeast lobe computed with MCNP compared to the generic PDQ factors.

The axial peak-to-average power factor is slower to change with burnup when the KJRR-FAI is present. This too can be attributed to the influence of the KJRR-FAI. It is noteworthy that because the KJRR-FAI is HALEU, as opposed to HEU, it has much more internal fertile-to-fissile conversion. This causes its own reactivity contribution to change slower with time. The KJRR-FAI generally drives the midplane power of the ATR fuel elements throughout the four cycles for which the test was irradiated. This caused an increase in the fuel burnup at mid-plane per assembly average burnup.

The combination of starting with heavily burned fuel elements to suppress lobe-power and the faster burnup rate of these elements required careful fuel element selection to ensure that the requested cycle-length could be achieved without exceeding the burnup limits on ATR fuel elements. Careful selection of fuel elements, essentially salvaging fuel elements slated for disposal, enabled the achievement of both the lobe-power and cycle-length constraints.

#### **3.3 Finding a new equivalency**

As mentioned previously, HELIOS assumes that all axial details are constant by nature of being a 2D code. This leaves the reactor analyst with one of two choices: extrude the most reactive axial region and assign this geometry and composition to the 2D HELIOS model (1), or axial homogenize all regions within the active core height and assign this composition to the 2D HELIOS model (2). For cases where it is important to preserve the overall reactivity worth of the experiment, its influence on lobe-power, and overall core reactivity, volume weighted axial homogenization is required. For cases where it is important to preserve the spatial self-shielding between the experiment and the nearby ATR fuel elements, extrusion is required. For KJRR-FAI, neither assumption could completely preserve the 3D behavior. Modeling the 21 fuel KJRR-FAI fuel plates as an extrusion artificially assigns the KJRR fuel density meant for 60 cm to the full 121.92 cm (48 in) active height of the ATR fuel. This would grossly over-estimate the fissile content of the test and artificially increase the reactivity contribution of the northeast lobe, thus producing a nonsensical estimate of required reactivity hold-down for the northeast OSCC quadrant. If the KJRR-FAI were homogenized with water and aluminum holders above and below it, the interplay between KJRR-FAI and ATR fuel element plates could be lost; thus, losing confidence in the burnup rate of the northeast lobe's fuel elements. The solution was a compromise between extrusion and homogenization.

By representing the KJRR-FAI mid-plane geometry in the HELIOS model, but reducing the uranium concentration in the fuel meat, the power of the test and its influence on power of the eight neighboring ATR fuel elements could be preserved. **Figure 10** shows the HELIOS computed KJRR-FAI power as a function of fractional uranium loading.

MCNP analysis showed that in order to keep the peak heat flux below the KJRR-FAI programmatic constraint (200 W/cm<sup>2</sup> ), the total fission power of the prototype fuel element would need to be kept to below 2.3 MW<sup>3</sup> . The minimum heat flux requested was 137 W/cm<sup>2</sup> ; thus, providing a lower bound of 1.6 MW. Therefore, the fractional uranium loading was reduced to 20% in order to provide a representative test power, as well as accurate power sharing behavior within the northeast lobe.

<sup>3</sup> The peak heat flux for the KJRR-FAI test program was 200 W/cm<sup>2</sup> . The test itself in the ATR northeast flux trap had significantly more thermal margin.

#### **Figure 10.**

*As-modelled power, using HELIOS, of the KJRR-FAI.*

#### **Figure 11.**

*Comparison of calculated fuel element power between MCNP with full 3D detail and HELIOS using a 2D model of the KJRR-FAI with 20% of the true uranium loading.*

Note that the KJRR-FAI fuel meat height is roughly half that of the ATR fuel element, yet the fractional loading is only 20%. This is intuitive if one considers that fuel near the midplane has a greater importance (i.e., considering flux-weighting) compared to fuel further from mid-plane.

The adjustment process is verified against MCNP calculation of the 158A core with OSCC in the startup position of 29 degree rotated-out. This is referred to as the startup power distribution. The comparison of ATR calculated fuel element powers is shown in **Figure 11**.

This adjustment process is validated by the fact that the northeast lobe-power tracks well when compared to the lobe-power measurement system. ATR uses the water activation reaction, 16O(1 n,p)16N ! 16N (*T*1/2 = 7.13 s) ! 16O + <sup>β</sup>� <sup>+</sup> <sup>γ</sup> to indicate lobe-power. Ten flow tubes, two at central, four at ordinal positions near the lobes, and four at cardinal positions beyond the OSCC provide activation information to ion chambers. The ten ion chamber signals are converted via the least squares method to compute lobe-powers by a monitoring computer. This computer *Core Reload Analysis Techniques in the Advanced Test Reactor DOI: http://dx.doi.org/10.5772/intechopen.103896*

#### **Figure 12.**

*Comparison of calculated versus measured (via the N-16 system) northeast lobe-power for ATR cycles: 158A, 158B, 160A, 160B.*

also records the gross calorimetric power of the reactor, as well as the OSCC, and neck-shim positions every hour. This information is combined into a post-cycle analysis of the ATR cycle using HELIOS. This "As-Run" calculation serves two purposes. It provides accurate fuel element depletion results which are then tracked for fuel management records. It also serves as a continuing improvement process for code maintenance of HELIOS and associated ATR models. A comparison of calculated by HELIOS versus measured by the N-16 system for all cycles containing the KJRR-FAI is shown in **Figure 12**. In the figure, the calculated northeast lobe power is shown with and without the KJRR-FAI. The KJRR-FAI power can be calculated by HELIOS, as was shown in **Figure 10**. However, calculating experiment power is not part of the typical As-Run process. Therefore, unfortunately, this data is not available. However, the MCNP As-Run for which provides data to the KJRR-FAI project is available and is included in **Figure 12**.
