Cardiac measures of nuclear power plant operator stress during simulated incident and accident scenarios

Abstract

Maintaining optimal performance in demanding situations is challenged by stressinduced alterations in performance. Here, we quantified the stress of nuclear power plant (NPP) operators (N 5 20) during a full-scale simulator training for incident and accident scenarios. We compared the ambulatory electrocardiography measurements of heart rate (HR) and heart rate variability (HRV), and self-reported stress during baselines and simulated scenarios. Perceived (scale 0-10) and physiologically measured stress were low during baseline after the scenarios and normal NPP operation (means 1.8-2.2, mean HR 75-80 bpm). During a cognitively challenging scenario simulating a sensor malfunction, the operators' stress was mild to moderate (mean 3.4; HR 1 12% from baseline). During simulations of severe accidents of fire and radioactive steam leakage, the experienced stress and cardiac activity were on a moderate to high level (means 4.2 and 4.6; HR 1 23% and 1 14% from baseline, respectively). Cardiac activity paralleled the self-reported stress: correlation of selfreported stress to HR was 0.61 (p < .001) and to HRV features RMSSD, HF, LF/HF, SD1, and SD1/SD2 were 20. 26, 20.28, 0.35, 20.40, and 20.39 (p < .01), respectively. The low shared variance (22%) between HR and physical activity further support the interpretation that the cardiac activity was strongly linked to the experience of stress and not accountable by operators' movement within the simulator. Cardiac measurements in naturalistic settings can thus reveal relevant information on acute stress with the benefit of not interrupting the primary task.

Modeling the cardiac indices of stress and performance of nuclear power plant operators during simulated fault scenarios

“…The very first period when the operators were filling questionnaires, before any of the incident and accident scenarios, was initially planned to represent baseline. However, as was seen already in the report published in Pakarinen et al, , and also here, the operators' experienced elevated levels of cardiac activity in the beginning of the experiment, interpreted as pretest excitement. Hence, the T5 and T3 with the lowest levels of stress were set as baseline.…”

“…Initially, the motivation for recording the operator movement was to estimate the effects of physical movement on cardiac activity. In our earlier work from this same simulator training (Pakarinen et al, ), the shared variance between the cardiac and physical activity and the experienced stress was low (22%), supporting the interpretation that the recorded changes in cardiac activity reflected mainly the stress physiology instead of operator movement. Also here, it is seen that the HR is strongly affected by the accident and incident events of the simulation training (Figures and ), whereas the physical activity remains similar between most events, and the strongest differences in the crews' movement are seen during the normal operation, which was the only event without a predefined structure, that is, either predetermined operations to perform or questionnaires to complete (Figure ).…”

“…The frequency domain features, for instance, are more strongly affected by respiration (for a review see, Laborde, Mosley, & Thayer, ), which is difficult to control in naturalistic experiments including free movement. Moreover, in our recent study (Pakarinen et al, ), with the same data as used here, a comparison between various HR and HRV features indicated that the mean HR was most strongly related to the stress of the operators. As we are interested in the effects of stress on operator crew performance, we continued along this line and went on to investigate the relationship between the median of the IBI (inverse of HR), and the operator crew performance.…”

“…During the most stressful event, that was, the power plant shutdown including both the reactor and the turbine shutdown, the crew HR increased on average by 16 bpm from the baseline level of 77 bpm. According to our earlier work examining the association of the cardiac activity and experienced stress, an increase of such magnitude corresponds to a moderate or slightly higher than moderate experienced stress (Pakarinen et al, 2018). Also, as reported earlier (Pakarinen et al, 2018), the operators were required to rate the highest level of stress they experienced during the accident and incident tasks on a scale from 0, denoting relaxed state with no stress at all, to 10 denoting extreme stress.…”

“…Quantifying stress via psychophysiological measurements has the benefits of providing a continuous time series that is free from response bias and obtained without interfering the primary task (see for instance Henelius, Hirvonen, Holm, Korpela, & Muller, 2009;Pakarinen, Korpela, Torniainen, Laarni, & Karvonen, 2018). One widely used method is to record the cardiac activity with electrocardiography (ECG) (Castaldo et al, 2015).…”