The 2022 Russian full-scale invasion of Ukraine has introduced many unprecedented and rarely-encountered challenges for the nuclear power generation and radiological safety communities. These include the occupation of the highly contaminated Chornobyl exclusion zone (ChEZ) by military force in February of 2022, the military capture and occupation of Zaporizhzhia nuclear power plant (ZNPP) in March 2022, and a drone strike on a the Chornobyl nuclear power plant New Safe Confinement structure in February 2025, to name a few of the most significant events. There have also been strikes on and near hospitals, mining facilities, and research centres that house nuclear material, fires on and near NPP facility grounds, and drone flights and strikes near other operating nuclear power plants (NPPs) in Ukraine and Russia that may either strike the NPPs themselves or cut off outside power and lead to a loss of coolant event.

All of these kinds of events can lead to different radiological release scenarios that have varying probabilities of occurrence and varying consequences for public health and safety. Despite the fact that military action has introduced these possible scenarios, it still falls to civilian agencies to assess how to monitor and prepare for them since the consequences of most scenarios can affect large areas and the public health and safety of a country’s population. Therefore, a robust radiological monitoring system should be used to autonomously measure radiation dose rates and telemeter information to a central monitoring database for adjudication of alerts and evaluation of data for decision-making. These data are also valuable for providing ground-truth to counter misinformation and disinformation.

In this paper we provide background on radiological emergency response preparation, considerations for implementing supplementary dose rate network based on current technology, and a case study for deploying a dose rate monitoring sensor network. Note that much of the technology and many techniques and strategies described in this paper are not highly original, but their application to radiation monitoring under wartime conditions is. This is intentional—systems deployed in such conditions need to be proven and resilient.

At the time of writing, there have been no large-scale radiological releases in Ukraine since the start of the full-scale Russian invasion in 2022. In February 2022, Russian occupation activities in the ChEZ resuspended radioactive contamination already present, but this event did not cause a radiological emergency [1]. In March 2022, Russian forces captured and began the occupation of ZNPP, which has introduced significant risk that military actions near the site will lead to a significant accident. These actions are what first led to the implementation of a supplementary dose rate sensor network in Ukraine because sensors in the ChEZ and at ZNPP went offline during the Russian occupation. The United States Department of Energy National Nuclear Security Administration (NNSA) deployed over 200 dose rate sensors to Ukraine to establish a supplementary dose rate monitoring network in preparation for any possible radiological release that might occur. This was done in cooperation and coordination with the government of Ukraine, including significant partnerships with the State Nuclear Regulatory Inspectorate and the Hydro meteorological Centre of the State Emergency Services. The supplementary network was established alongside the existing network to provide additional coverage, resiliency, and confidence in case the primary monitoring network fails, e.g. due to munitions strikes, power outages, electronic interference, or cyberwarfare. The initial goal of the deployment was to monitor for any kind of radiological release that occurred anywhere in Ukraine, at any time. This goal supported both health and safety decision-making for Ukraine as well as desired U.S. situation awareness. The supplementary network operated as a complement to the existing network of Ukrainian sensors, mainly located around NPPs and in the ChEZ, which experienced issues during the first Russian attacks.

The broad goal of comprehensive monitoring was addressed in several ways, including by working with Ukrainian partners to deploy dose rate sensors near Ukrainian NPPs, in the ChEZ, and in population centres across the country. Because of the emergency nature of the network, it was implemented in stages as resources became available. The first shipment included only sensors immediately available through interagency U.S. resources. Subsequent sensor shipments were sent in batches as they were procured and produced to build the capability as it became available, instead of delaying the deployment until the entire capability was ready and potentially being unprepared for a radiological release. Eventually the network was considered sufficient given the coverage and practical limitations. These sensors operated on battery to avoid reliance on the power grid, which at one time was the primary target of Russian strikes, and used satellite communications to avoid reliance on cellular or wired networks, which were also subject to power interruptions. Operation on battery power also required that batteries be swapped regularly to ensure operation of the sensors. This was accomplished by in-country implementers contracted through NNSA.

As the war in Ukraine continues, the need for the monitoring network remains. Despite the fact that the likelihood of a significant radiological release has not decreased, the emergency pace of operations cannot be sustained, so operations have shifted into long-term preparation and monitoring. The initial situation in 2022 can be likened to a sprint where significant resources were marshalled to quickly prepare a nation-wide supplemental monitoring network. However, the current situation is more akin to a marathon where resources need to be carefully considered and allocated to best prepare for long-term radiological monitoring. This change in pace prompted an assessment of the emergency solution implemented starting in 2022 and whether that solution needed to be changed to account for sustained monitoring operations. The goal of the supplementary dose rate network, an important factor for deciding on instrument requirements and deployment strategy [2], was reconsidered in the context of long-term monitoring using elements of the data quality objectives process [3], namely consideration of the main questions to be answered using data collected by the network, the budget and schedule for implementation, the data needed, and how alarms would be adjudicated. The initial deployment assumed sensors might be deployed for a matter of months, but as the initial military actions have become a more prolonged war, it is necessary to re-assess the state of radiological monitoring to determine how much of the original and supplementary sensor networks need to change—increasing or decreasing the number of sensors, changing sensor hardware for longer-term sustainability, and changing deployment configuration (e.g. locations) to better suit long-term goals. An example of this re-assessment and planning process is provided in section 4.

When considering the technology to employ in a wartime emergency scenario, there are many factors to consider. Automated, ‘headless’ sensors were the baseline considered for use in this scenario. Using such sensors means that personnel are not needed to remain in an area and collect measurements on a regular basis, which reduces the labor-intensiveness of the network and, more importantly, reduces personnel exposure to wartime hazards. There are four main components that make such a system operational: the radiation sensor package, data telemetry backbone and services, a database to collect the telemetered data, and an analysis process and experts to evaluate the data. Finally, the advantages and disadvantages of networks with heterogeneous and homogenous sensor models are discussed.

3.1. Radiation measurement sensors

For the purposes of this discussion, radiation measurement sensor package can be described as having three elements: the sensing element and associated electronics, the power source, and the data transmission module. The first two elements are discussed in this section, while the data telemetry is discussed in section 3.2.

Dose rate sensors are typically simple, mechanically and electronically. They are typically composed of single or dual Geiger–Mueller tubes, which are a mature technology based on well-understood and simple physics principles, largely insensitive to environmental temperature fluctuations, mechanically robust, and low-cost in comparison to other ionising radiation measurement technologies, such as gamma-ray spectroscopy systems. Geiger–Mueller-based dose rate sensors are therefore an ideal type of radiological sensor for the application discussed here—a large, geographically dispersed sensor network for the purpose of providing an initial verification of a radiological release. Dual Geiger–Mueller tubes are often used to cover an extended dose rate range compared to a single tube, which is important for our application since the radiological environment may change rapidly and significantly over a short time period.

One significant limitation of dose rate sensors is the inability to provide data that can be used to identify and quantify radionuclides and therefore distinguish changes in natural background from the presence of potentially harmful releases of radionuclides. Such capability is important for accurate internal dose assessment (i.e. for inhalation and ingestion pathways) and is typically supported by spectroscopic sensors. While dose rate sensors are more cost-effective, it may be advantageous to consider adding spectroscopic sensors to a network. The value of adding such systems to a network depends on the resources available—for example, the funding available and whether such systems are likely to be present already in the hands of first responders.

Including an independent power source for the dose rate sensor networks is also important. Since the purpose of a supplementary network is to continue operating even when existing networks fail. Batteries are an easy and ubiquitous way to power dose rate sensors independently, since the sensors themselves typically require little power. If the batteries required are small enough, the whole sensors can be made small and light enough to be repositioned easily, another advantage for an emergency network. In addition to batteries, solar panels can help provide additional wall power independence, depending on expected sunshine at installed locations.

A final consideration that should be made is legal defensibility of the sensor data. Depending on local regulations, sensors may be required to undergo certification to be permitted as measurement instruments whose data can be used to inform official decisions. While it may not be required in the emergency phase of operations, as the emergency phase transitions into longer-term intermediate phase operations, certification should be considered, if applicable, to impart stronger reliability and acceptance of sensor data. This can become especially important if false narratives and misinformation about radiological releases can arise. Reliable, certified sensors and data can be strong ways to provide ground truth in such situations. Uncertified sensors may also provide useful data that should be considered for adjudication and informing emergency response, especially if no other data is available in an emergency or conflict environment. However, despite extenuating circumstances, data from uncertified sensors may not be legally used for decision-making in some cases. The authors recommend that regulatory frameworks are flexible enough to allow the use of uncertified sensor data in emergencies.

3.2. Data telemetry

For any sensor network, the sensors need to telemeter their measurement readings to a database for analysts to evaluate. Selection criteria for sensor communications are at least equal in importance as the type of sensor chosen—without a way to collect the data, the sensor is useless. Telemetry also serves as a method for confirming that the sensor is still operational. When selecting telemetry options such as Ethernet, Wi-Fi, cellular, and satellite, there are several considerations.

The first is to consider how much data needs to be sent with each transmission, in terms of bytes. Dose rate sensors transmit not only a dose rate numerical value, but also metadata including measurement unit, date, time, latitude, longitude, identification number, battery capacity, and voltage. These data can easily be encompassed within the megabits per second range of most modern communications options.

A related but separate consideration is how frequently to transmit measurements and metadata. If there is no emergency, measurements do not need to be transmitted frequently because the data are not changing. If there is a radiological release then organisations receiving measurements will want to obtain data more quickly to provide timely public health and safety guidance. More frequent data transmission is in opposition, however, to power consumption and cost of the system. The power needed for a sensor communications module is often significantly greater than the power needed by the sensing module itself. The more often data are transmitted, the more power the sensor will need (which is a major consideration if the sensors are battery or solar-powered, for example) and the greater the cost of the communications services will be. Striking the proper balance is a matter of assessing how quickly information will be needed in an emergency situation versus the resources available to the organisation deploying the sensor network. If possible, it is advantageous to configure sensors to measure and transmit at a ‘slow’ rate when measuring normal background levels, and measure and transmit at a ‘high’ rate if a dose rate value above a certain threshold is measured. For example, an alert level of 400 nSv h−1 could be configured such that, when measurement values are below 400 nSv h−1, the sensors measure every 60 min and transmit every 24 h, and when measurement values are at or above 400 nSv h−1, the sensors measure every minute and transmit every hour. This scheme preserves battery capacity when in non-emergency conditions, but prioritises data collection when there are elevated dose rate levels. The alert level of 400 nSv h−1 is used here as an example only. Organisations that wish to employ such a scheme should determine thresholds appropriate for their specific public protection thresholds, goals, situation, and background dose rate.

Another factor to consider is what communications pathways can be relied upon, especially given a wartime environment. Battlefield operations increasingly feature electronic warfare, the effects of which can extend beyond the active battlefield. Such effects can include global positioning system (GPS) jamming which can also affect data from deployable dose rate sensors. In other situations, loss of local power can mean that Ethernet/WiFi and cellular are interrupted. It is best, therefore, to include at least two communications pathways that share few or no failure modes between them for resiliency. Implementing multiple communication pathways is also at odds when trying to reduce power consumption and cost of the sensor. Even if both communication pathways are not used at the same time, there is usually residual power consumed by the inactive pathway, and switching between pathways requires power. Including more than one communication pathway also means the capital cost of the sensors increase because of added engineering costs needed to make sure all pathways work and that the sensor will switch between them automatically. The communication cost of the sensors may also increase because more than one service will need to be active constantly to facilitate automatic pathway switching.

3.3. Measurement database

The sensors will need to telemeter their data to a digital location where subject matter experts can view and analyse data. A database is recommended to best organise transmitted data though data can also be collected and stored in other ways. Such a database will need to include a method for adding new data automatically from each of the sensors, as well as extracting or viewing data upon request. Such a database should be able to receive measurements transmitted at any time and arrange the measurements according to the time measured for ease of analysis. For example, a sensor may transmit measurements collected most recently first, and then transmit older measurements. However, these measurements should be arranged least to most recent in the database to aid time-series analysis data. As discussed in the previous section, the sensors ideally also transmit metadata such as the date and time the measurement was collected, the location, battery status, and any other available data, such as temperature. The database should therefore also be able to display these data as they correspond to each measurement or to the device as a whole. An example view of the database information is provided in table 1.

Table 1. Example of data fields recommended for inclusion in a database collecting automated dose rate measurements, including the dose rate measurement values, units, and corresponding metadata.

Device IDMeasurement dateDatabase collected dateValueUnitLongitudeLatitudeBattery voltageDosimeter 11 January 2026 12:50:001 January 2026 13:00:00105nSv h−138.8874−77.025912.0 VDosimeter 11 January 2026 12:40:001 January 2026 13:00:00103nSv h−138.8874−77.025912.0 VDosimeter 11 January 2026 12:30:001 January 2026 13:00:00108nSv h−138.8874−77.025911.9 VDosimeter 11 January 2026 12:20:001 January 2026 13:00:00110nSv h−138.8874−77.025911.9 VDosimeter 11 January 2026 12:10:001 January 2026 13:00:00104nSv h−138.8874−77.025911.9 VDosimeter 11 January 2026 12:00:001 January 2026 13:00:00106nSv h−138.8874−77.025911.9 V

Except in cases where the sensor network is homogeneous, i.e. sensors are all one make and model, the database will often be separate from data collection applications to which sensors first telemeter their data. This is a purely practical consideration since commercial sensors are often configured to report data directly to a manufacturer’s proprietary software using a proprietary communications format or file. This often means that it is easier to develop the data pathway such that the proprietary software then sends the data onward to the measurement database. In some cases, the ability to send measurement data to an email address or separate cloud-based database may already be built into a manufacturer’s proprietary software.

The database ideally also includes a graphical user interface that displays the data in a number of different ways, including a map view and lists of measurements collected arranged by time the measurement was collected. When analysts look at the data, they will need to quickly view data from multiple sensors, compare them, and compare to other data, such as precipitation, wind velocity, and temperature. They will also need to be able to quickly search for sensors, potentially from among hundreds of different serial numbers, and locate them on a map.

To perform detailed analysis, it is also useful for the database to have a way for analysts to extract data to a local computer. This can be used for detailed analysis (more than the ‘at a glance’ analysis that can be conducted in the database GUI) and to automatically generate custom data products, such as a report on the network state of health. One extraction method is an application programming interface (API). APIs are common for many applications that deal with downloading data from a website or allowing two tools to exchange data so that users can seamlessly work between both tools and reduce data handling errors that come with manual, human data handling.

Data ownership should be addressed in the early stage of system design so data can be shared appropriately and quickly when needed. Typically, the country that owns the sensor owns the data. This is important when the theoretical dose rate sensor network discussed in this paper is implemented in partnership between the country involved in military conflict and those providing support. When a supporting country provides sensors, the sensors may remain property of the supporting country, or they may be donated and ownership transferred to the country involved in the conflict. The decision about which path to follow is outside the scope of this paper. However, in both cases, both countries will need to agree on data ownership and who will receive data, including possible third-party partners.

3.4. Analysis

A dose rate network like that described here will require analysts and a standard analysis procedure to evaluate and adjudicate dose rate sensor data to produce actionable recommendations for public safety.

First, monitoring a dose rate network should not be conducted by a human analyst continuously. Even for a small network, there is simply too much data that, most of the time, is not changing significantly. A more efficient method for monitoring dose rate sensor network data are to implement automatic algorithms that alert the right personnel to potential issues based on certain conditions. These algorithms can be run either as part of the database described in section 3.3, or separately through another service that is accessing data from the database. Such algorithms could include simple threshold-crossing based on a ‘sigma-above-background’ value, or any number of more sophisticated methods, including the use of artificial intelligence. The discussion of such methods is beyond the scope of this paper—it is enough to state here that some kind of automated algorithm should be implemented to alert human adjudicators to review the data. An algorithm or artificial intelligence can maintain uninterrupted awareness for long periods of time, helping human adjudicators focus on alerts only when needed.

Analysts should be ready to evaluate and adjudicate alerts from the network at any time of day, even outside of wartime conditions. Therefore, a cadre of on-call analysts needs to be established to be the first to view data when automatically alerted, and who can potentially alert others if it appears that a true emergency is underway. How to properly establish such a cadre, the concept of operations, and the evaluation and adjudication process are beyond the scope of this paper.

3.5. Network uniformity

As described in section 2, deployment of a supplementary dose rate sensor network in an emergency situation, especially in wartime conditions, is unlikely to happen all at once. It is expensive to maintain a large collection of dose rate sensors stored for eventual use, and therefore likely that sensors will be deployed in phases toward prioritised objectives. Even if one deployment is all that is required for a particular case, sensors on-hand may be of all one make and model, or may be heterogeneous. There are certain advantages and disadvantages to each configuration, and those advantages and disadvantages are mutually exclusive, i.e. an advantage of one configuration is a disadvantage of the other. These advantages are summarised in table 2.

Table 2. Dose rate sensor network parameters and whether they constitute an advantage for networks of homogeneous or heterogeneous sensors.

ParameterAdvantage forHomogeneousHeterogeneousPurchasing—number of manufacturers, points of contact, ordersX Software subscriptionsX Number of different spare parts to trackX Maintenance implementer experienceX Data integration—number of different data protocols neededX Sensor settings—number of different configurations to trackX Cyber resilience—number of different systems to compromise XMaintenance issues—number of sensors that may experience the same maintenance issue XProduction speed—number of sensors that can be manufactured in parallel XSupply chain resilience—number of sensors that rely on the same parts bottleneck X

The main advantage of a homogeneous sensor network is efficiency. An ideal homogeneous sensor network uses a single make and model of sensor, and needs only a single equipment purchase, manufacturer point of contact, software subscription, spare parts set, data integration plan, and sensor settings configuration. In addition, deployment and maintenance efficiency is also improved since the implementer only needs to learn the process and procedures for one sensor manufacturer and model. The specific situation can become more complicated, e.g. if another batch of the same make and model of sensor is needed to replace lost or destroyed sensors. But generally a homogeneous network is more efficient than a heterogeneous one because the complexity (and therefore cost) only increases if more sensor makes and models are added to the network. Adding even one more sensor make and model doubles the number of purchase orders, points of contact, software subscriptions, data integrations, and sets of settings to track.

On the other hand, the main advantage of a heterogeneous sensor network is resilience. Using sensors with varying makes and models means that the network will be more resilient to cyberwarfare since each different sensor uses a different set of security measures—compromising one sensor does not compromise the whole network. A heterogeneous network is also resilient to maintenance issues. If an issue is discovered with one sensor, e.g. it was not properly sealed against moisture, it is unlikely that all sensors in the network will have this issue. Finally, using a heterogeneous network adds resilience to manufacturing bottlenecks, e.g. supply chain issues or factory throughput. Though radiation sensors like Geiger-Mueller tubes are ubiquitous, radiation sensor manufacturers typically do not keep a large stock of equipment ready for immediate shipment because the market for such equipment is relatively small. Therefore, when procuring sensors for a large network, it may not be possible to have enough sensors immediately available from a single manufacturer. Of course, if different manufacturers still rely on a single supplier for a key part, then diversifying manufacturers in the network may not help. But diversifying manufacturers does reduce the likelihood of such a circumstance occurring.

The authors do not make a recommendation whether a network should focus on efficiency or resiliency. Using a homogeneous or heterogeneous network is a matter of the network goals and operational circumstances of each unique emergency situation.

When considering implementation of a dose rate sensor network to monitor for radiological releases, the most important factors are (1) establishing clear, achievable goals against which implementation can be measured, (2) how many dose rate sensors are required to achieve the system goals, and (3) where do the sensors need to be placed to achieve the system goals, and (4) what requirements dose rate sensors must meet to be suitable for use in the supplementary network.

4.1. System goals

The first version of an emergency-phase goal of the dose rate sensor system might be to monitor for any kind of radiological release that occurred anywhere, at any time. Truly accomplishing this goal, however, is not practical because it would require thousands of sensors distributed evenly across the country and cost tens to hundreds of millions of dollars (USD). Even such a deployment might not detect every incident depending on the magnitude of the release and specific conditions at the time.

Therefore, it is better to set a more practical goal. The scope of the supplementary network should thus limited to monitoring for releases based on the consequences, probability, or the product of the two (risk). How to quantify the consequences and probability of different release scenarios varies and is not discussed here, but as assessment should be made and a threshold for concern should be established and the chosen metric (consequence, probability, or risk) compared to the threshold to decide whether a specific facility type or specific facility itself should be covered. However, if the specific probabilities of scenarios cannot be quantified to sufficient precision, it may be better to choose to monitor for the greatest consequence scenarios, i.e. NPP and radiological waste storage sites, as depicted in figure 1. We will assume such a case for simplicity in the remainder of the example herein. Therefore, the goal of the example system will be to monitor for releases from NPPs and radiological waste storage sites. In this case, we define ‘monitoring for releases’ to include (A) promptly (within minutes) detecting that a release has happened, (B) indicating the direction(s) in which the release is moving, and (C) monitoring dose rate from the release in large population centres to inform public health and safety decisions. This level of specificity is helpful when constraining design to a practical system, in part because it is much different than other radiation detector network applications [4]. Because of the emphasis on protecting populations and the fact of limited resources, it may be necessary to limit the number of population centres deployed against to those locations with the largest populations, often cities or large metropolitan areas. It is left to planners and decision makers how to choose and prioritise these locations.

Figure 1. Consequence and probability of different radiological release types. Not all possible types are shown.

Download figure:

Standard image High-resolution image 4.2. Sensor placement strategy

An example of sensor placement strategy is presented here based on monitoring NPPs. The strategy comprises three sets of detectors, depicted in figure 2: (a) a ‘ring’ of dose rate sensors surrounding the NPP (‘ring set’), (b) a set of ‘interceptor’ sensors between the NPP and nearest large population centres (‘interceptor set’), and (c) a set of sensors arranged in an arc around a city in the direction of the NPP (‘city set’). These three sets of sensors correspond to the three specific goals, A, B, and C, described in section 4.1, respectively. A key parameter to consider is the maximum distance between population centres and NPPs at which a population centre is considered ‘in the path’ of a potential plume release. This will ultimately dictate how many population centres total are to be considered. How to settle on a quantitative distance can be accomplished in many ways, but should take into account the facts that, if it is too short then major population centres may be missed, and if it is too long then resources may be stretched too thin.

Figure 2. Diagram of the dose rate sensor placement strategy for an NPP.

Download figure:

Standard image High-resolution image

How many sensors need to be deployed in each set must be determined next. For the NPP ring, atmospheric transport simulations using standard NNSA atmospheric dispersion modelling tools [5] were conducted to determine how much a release plume would spread, based on diffusion and movement of air, and the dose rate at ground level that would result. These simulations were performed for a general, worst-case NPP release and did not include geographic considerations in order to make them widely applicable. The simulations were not intended to address a specific situation, but rather representative of the conditions under which the greatest amount of radioactive contamination would be spread the greatest amount. This was compared to the minimum dose rate sensitivity of common dose rate sensors available, typically on the order of 10 nSv h−1, to determine how far apart around the virtual ‘ring’ sensors could be spaced without detecting a ‘typical’ plume. At a distance of 10 km around the plant, it was determined that 19 dose rate sensors were needed for the ring. This analysis is also specific to the source term expected from an release scenario, so results may vary based on the specific reactor model considered.

For the interceptor set, atmospheric transport calculations were also used to assess the spread of a plume following an NPP release. The results showed that, even in the case of minimum realistic plume spread, a given plume will be wide enough at a range of tens of kilometres that a single dose rate sensor is needed as an interceptor between a single NPP and a single population centre. This assessment may change for population centres in the potential path of plumes from multiple NPPs, especially if they are in different directions from the city, and if the population centre is more spread out geographically. For example, if a single population centre is in the potential path of two NPP releases, then at least two sensors should be allocated as interceptors and one each located between the population centre and each NPP.

The number of sensors needed for the city set may vary based on several different factors. Two major considerations include the geographical size of the population centre to be protected and the dose rate exposure magnitude of the expected plume and deposition. It is left to planners and decision makers how to establish how many sensors are needed for a specific case, but it is recommended that a metric be developed to help guide proportional sensor allocation. One metric that establishes a rank for each population centre can be calculated quickly is provided as equation (1):

where P is the power of the NPP being considered (e.g. in MWe), ρPOP is the population density of the population centre (e.g. in number of people per km2), D is the distance between the NPP and the population centre, and L is an optional scaling factor to make the ranking metric more easily understood by human analysis. This ranking metric increases with physical population centre area and reactor generating power, and decreases with population size and distance from the reactor. This metric was developed such that a greater size of the area that needs to be monitored and a greater amount of material that could be released (roughly proportional to reactor power) increase the ranking metric value, and a greater number of people in the area (and thus other resources) and greater dilution of the release concentration (proportional to distance from reactor) decrease the ranking metric value. An example of fictional NPPs and population centres, and their corresponding rankings are presented in table 3. The population centres, A–E, range in population from 100 000 to 1000 000 people, range in area from 1 to 1000 km2, and range in distance from 50 to 250 km from the NPPs of concern. Two NPPs are presented, NPP1 and NPP2, with power output of 3500 and 2000 MWe, respectively. The resulting metrics range from 3.5 to 35. It could, for example, be decided that two sensors will be allocated to city sets with a rank 15 and above, and one sensor to city sets with a rank below 15. This would mean that population centres A, D, and E would all be allocated two sensors. Note that, in the table, only the maximum power of all reactors within 250 km is considered. Alternate ways to implement this ranking method would be to sum the power output of all NPPs within 250 km, or simply to allocate an additional sensor for each city set for each additional NPP within 250 km beyond the first.

Table 3. Table of fictional population centres and NPPs used as an example for calculating ranking metrics.

Population centrePopulationArea (km2)NPPs within 250 kmMax P (MWe)D (km)LRanking metricA100 00010NPP135001001.0 × 10−0435.0B200 0001NPP13500501.0 × 10−043.5C500 00020NPP1, NPP235002001.0 × 10−047D700 000300NPP220002501.0 × 10−0434.3E1000 0001000NPP220001501.0 × 10−04133.3

Depending on resources available, it may not be possible to deploy the number of dose rate sensors suggested by the example methods above, or other methods developed by individual practitioners for their specific situation. In such cases, prioritisation is helpful to decide which parts of the complete plan are most important and should therefore be allocated the limited resources. As an example, in the method above, the complete plan calls for dose rate sensors deployed near NPPs, at interceptor locations, and in population centres.

If resources are limited, elements of the plan can be prioritised in two ways. First, the population centres can be prioritised by score and possibly only allocated one sensor to reduce the resources needed. Second, some locations may not need automated dose rate sensors if it is determined that sufficient timely data will be collected via other means, such as personal radiation detectors, handheld dosimeters, or other equipment expected to be carried by first responders. Such equipment is common in large population centres, so it may be determined that automated, supplementary dose rate sensors are not needed in large population centres. However, smaller population centres may not have such equipment, and deployments may be warranted in such areas.

4.3. Dose rate sensor specification requirements

A set of requirements should be developed against which detectors are compared to be considered suitable for use in the emergency supplementary network. It is important to consider only the immediate, technical requirements for use, and not regulatory requirements that might normally be imposed on such sensors in peacetime. It is most important to have some reliable data in an emergency, and it may not be able to meet regulatory requirements, such as certification for use as official metrology equipment, under emergency conditions.

Recommended technical requirement parameters to consider include:

1.  

Minimum and maximum measurable dose rate

2.  

Environmental conditions (e.g. temperature, moisture and dust ingress protection)

3.  

Power source and power longevity

4.  

Data telemetry method (e.g. cellular, satellite)

5.  

Measurement and telemetry frequency

6.  

Sustainment capability by the country hosting the sensors

7.  

Equipment, communications, and maintenance costs

For each parameter, it is recommended that both a threshold and objective requirement be set where possible and appropriate. The threshold acts as the absolute acceptable criterion for a particular sensor to be considered, and an objective acts as the criterion for distinguishing preferred sensors from others. An example of a table of thresholds and objectives for a set of sensor requirement parameters is shown in table 4. To evaluate sensors against these requirements, the empirical or manufacturer-provided instrument performance metrics for the sensors must be collected. An example method for performing such an evaluation is to assign a score to each metric for each candidate instrument based on whether the instrument meets the threshold (score = 1), the threshold and objective (score = 2), or does not meet the threshold (score = 0). The parameter score might also be weighted by multiplying the number of parameters affected in table 4, column 2, by the score received for that parameter, for example. The weighted scores for each parameter are then summed to a total weighted score.

Table 4. Table of example requirement parameters for dose rate sensors, with example thresholds and objectives.

RequirementParameter affectedBenefitThresholdObjectiveDose rate (low)Radiological, operationalSensitivity50 nSv hr−120 nSv hr−1Dose rate (high)Radiological, operationalHealth and safety1 Sv hr−110 Sv hr−1Temperature rangeOperationalYear-round operability−20 °C to + 40 °C−30 °C to + 40 °CWater and dust protectionOperationalWeather survivabilityIP 67IP 68Measurement intervalRadiological, operational, sustainabilityRapid response to changes60 min10 minReporting intervalOperationalRapid reporting of changes60 min10 minBackup power durationOperational, sustainabilityImproved maintenance cycle, resilienceBattery power providing 72 h operation Solar chargingOperational, sustainabilityImproved maintenance cycle, resilienceSolar charging option available Telemetry methodOperational, sustainabilityResilience, flexibilityCellularCellular and satellite backupReport positionRadiological, operationalTrack anomalous readings, sensor movementReport GPS position hourly 

Table 5 shows three hypothetical candidate instruments and their corresponding example values. These are only hypothetical instruments and though the values are based on realistic possible instrument performance values, the specific combinations of values do not correspond to any specific commercial instrument. Table 6 shows an example of a set of scores for the hypothetical candidate instruments listed in table 5, the weight assigned to each requirement, and the total weighted scores. The total weighted score can be compared to the maximum total weighted score possible to assess how well a given instrument meets the requirements. Note that, for requirements where only a threshold is present, the maximum unweighted score for that requirement is 1. Therefore, the maximum weighted score in the example in table 6 is 31.

Table 5. Table of parameters for three hypothetical dose rate sensors.

RequirementCandidate A valuesCandidate B valuesCandidate C valuesDose rate (low)20 nSv h−110 nSv h−120 nSv h−1Dose rate (high)