A 38 North exclusive with analysis by Joseph S. Bermudez Jr., Mike Eley, Jack Liu and Frank V. Pabian.
Thermal imagery analysis of the Yongbyon Nuclear Scientific Research Center indicates that from September 2016 through June 2017:
- The Radiochemical Laboratory operated intermittently and there have apparently been at least two unreported reprocessing campaigns to produce an undetermined amount of plutonium that can further increase North Korea’s nuclear weapons stockpile. This suggests batch rather than continuous processing of spent fuel rods from the 5 MWe Reactor during the period of analysis.
- Increased thermal activity was noted at the Uranium Enrichment Facility. It is unclear if this was the result of centrifuge operations or maintenance operations. Centrifuge operations would increase the North’s enriched uranium inventory; however, based on imagery alone, it is not possible to conclude whether the plant is producing low or highly enriched uranium.
- The thermal patterns at the probable Isotope/Tritium Production Facility have remained consistent, suggesting that the facility is not operational, or is operating at a very low level. This means, the facility is likely not producing tritium, which is an essential isotope used in the production of boosted yield nuclear weapons and hydrogen bombs.
- From December 2016 through January 2017, the thermal pattern over the Experimental Light Water Reactor (ELWR) was elevated. While that might indicate that the reactor was operational, the likelihood is low since the pattern does not appear in subsequent imagery over the last six months. It is possible that there are alternative explanations for the elevated pattern, for example, short-term activity at the ELWR such as the heating of pipes to prevent freezing. Regardless, any activity at the ELWR is cause for concern and bears continued monitoring.
- The 5 MWe Reactor has either been intermittently operating at a low-level or not operating. The notable exception to this was during December 2016 and January 2017 when thermal patterns suggests a higher level of operations.
While commercial satellite imagery is now widely used to analyze important developments overseas, including in North Korea, thermal imagery can provide additional important insights. Landsat 7 imagery from September 2016 through June 2017 was used for this analysis, although heavy cloud cover precluded the use of imagery from last November and no night-time imagery was available for the entire time period of this study. A total of 19 images are available and of these, 10 were chosen with approximately one-month time intervals between them to provide a consistent periodicity for the analysis. Seven images were deemed too cloudy for analysis and thus weren’t considered.
Developments noted at key Yongbyon installations were as follows:
Radiochemical Laboratory: Examination of the thermal patterns associated with the Radiochemical Laboratory (reprocessing facility) show significant deviations from month to month. Concentrated heat patterns were observed with stronger temperature differences from the surrounding area between September to October of last year. The thermal patterns then returned to lower levels until March 2017, when a distinct increase in thermal activity is observed that has continued through last month. These intermittent surges in thermal activity suggest North Korea has conducted batch rather than continuous processing of spent fuel rods from the 5 MWe Reactor. It is typical to allow the spent fuel rods to rest for a while in cooling ponds to both cool and allow less stable plutonium isotopes (PU-238, etc.) to bleed off. These reprocessing campaigns do not necessarily occur immediately after spent fuel rods are removed from the 5 MWe reactor. The June 2017 thermal activity coincides with an increase in activity noted in a March 2017 analysis based upon natural color imagery.
Uranium Enrichment Plant: The thermal patterns at the Uranium Enrichment Facility were elevated during September and October 2016, then decreased in November 2016 and remained low until March 2017 when it increased slightly. It is unclear if the period of elevated activity from September through November was related to centrifuge operations or the maintenance activity that was observed during this period.
Experimental Light Water Reactor: The same elevated thermal patterns over the 5 MWe Reactor observed in imagery during December 2016 and January 2017 also extended over the area of the ELWR. This was likely the result of steam being released into the air when the turbines adjacent to the 5 MWe Reactor were being run, operation of the 5 MWe Reactor itself, mid-winter heating of both structures, prevailing weather patterns, or some combination of the above. We cannot completely, however, eliminate the possibility that this elevated thermal pattern was the result of short-term activity at the ELWR itself—for example, heating the structure to prevent pipes from freezing, allowing ongoing internal construction work, or pre-startup testing. It is important to note that no other significant patterns of thermal activity were observed over the ELWR throughout the study period. Importantly, the ELWR did not operate at all from February through June 2017. Any activity at the ELWR is cause for concern and its operational status bears continued monitoring as it would be an indicator of North Korean ongoing intentions and capabilities.
5 MWe Reactor: The thermal patterns observed at the 5 MWe Reactor remain relatively consistent with those observed in the previous report indicating either intermittent low-level or no operation of the reactor. There was a notable deviation in the December 2016 and January 2017 images, suggesting a period of higher level reactor operation that lends support to a previous analysis based upon natural color imagery.
Isotope/Tritium Production Facility: The thermal patterns at the probable Isotope/Tritium Production Facility have remained consistently low throughout the period under study, suggesting that the facility is not operational, or is operating at a very low level.
Figure 1. Overview of the 5 MWe Reactor, ELWR and Radiochemical Laboratory (reprocessing facility).
Figure 2. Thermal patterns over these areas from October 2016 through June 2017. (Scroll through slide show.)
Figure 3. Overview of the Uranium Enrichment Facility and probable Isotope/Tritium Production Facility.
Figure 4. Thermal patterns over these areas from September 2016 through June 2017. (Scroll through slideshow.)
Landsat 7 imagery was used exclusively due to an ongoing issue causing reduced accuracy of Landsat 8 thermal sensors.
It is important to understand several limitations when using Landsat 7 thermal imagery for this type of analysis. Among these is that the data is collected at 60 meters ground sampling distance (GSD). That is, each pixel represents 60 meters. It is then resampled down to 30 meters GSD. Additionally, the Landsat 7 thermal sensor is measuring both the thermal response of the ground plus the air column above it. It is not unusual for a prevailing wind, which over Yongbyon at this time of year, generally originates out of the northwest or north, to carry the heated air column from one point to another (e.g., the 120-150 meters from the 5 MWe Reactor over the ELWR). In the current and previous 38 North thermal studies, we are pushing Landsat data to its limits.
Typically, a reactor would be loaded and started up at low power to heat and pressurize the coolant and adjustments would be undertaken to balance the pressure and flow through the system. Once operators are assured that the settings are correct the system can then be operated at full power. Once reactor operations commence, there is a requirement to continuously cool it to remove heat, even if it is only from radioactive decay. If a reactor has started up and subsequently shut down after a short period of low-level operation, the residual heat removal system would not need to remove significant heat because the core would be new. That heat is carried away by cooling water and never gets into the air.