**1. Introduction**

The corona pandemic in 2020 affects society, as well as essential services such as the water utilities, in a new and profound way [1]. Critical infrastructures like water utilities and the provision of their vital services in such a scenario have an outstanding importance to a nation's society. Their failure or degradation could result in sustained supply shortages, which affect public health, economy and national security. Quarantined personnel, working from home to distance employees and unpredictable supply chains for consumables are new and unfamiliar conditions that make the operation of water supply companies more difficult and potentially endangers overall water supply [2].

A resilient drinking water supply is consequently one of the basic requirements for a stable social and economic system. However, impairments cannot be completely avoided, so that water supply companies have to deviate from normal operation, e.g., in the event of pipe bursts. Such minor disturbances occur comparatively frequently and have only minor effects [3]. They can usually be quickly identified and repaired. As a consequence they usually remain unnoticed to the consumers [4]. On the contrary, failures or more extensive impairments of the water supply systems can have serious impacts on the affected population and the economy [3,5–7]. Causes can be serious natural events, man-made accidents or intentional attacks [8,9], whose probability of occurrence is constantly increasing [10].

The corona pandemic puts the importance of critical infrastructures, and the need for proactive emergency preparedness planning to increase their resilience, at the forefront of civil protection and disaster managemen<sup>t</sup> [1]. The understanding, analysis and quantification of resilience by water utilities, authorities, decision-makers and other stakeholders is a prerequisite for this.

The resilience of water supply systems can be increased by appropriate emergency preparedness planning instead of ad hoc coping responses. This includes the conceptual, organisational and technical preconditions for risk reduction and prepares structures for response in the event of a crisis [11]. Emergency preparedness planning in the water supply sector thus comprises, in addition to measures to avoid damaging events, especially preventive, safeguarding, reactive and restorative aspects of risk and crisis management.

Effective emergency preparedness planning is characterised, among other things, by the fact that the planning is carried out as preventive measures and the measures can be implemented in emergency situations. Beyond preventive measures to minimise risks, emergency preparedness planning in the water supply sector includes in particular aspects of crisis managemen<sup>t</sup> [12]. Such emergency preparedness planning takes into account different scenarios and their possible effects on the water supply. In addition to preventive measures, the numerous aspects of crisis managemen<sup>t</sup> also lead to risk minimization by limiting the extent of damage. Figure 1 shows the five steps of risk and crisis managemen<sup>t</sup> according to the German Federal Ministry of the Interior [13] and the Federal Office of Civil Protection and Disaster Assistance [12,14].

Thorough preliminary planning forms as the first process step the basis for the successful establishment of risk and crisis managemen<sup>t</sup> [13]. Basic specifications should be made in advance of the establishment or expansion of a risk and crisis managemen<sup>t</sup> system. These include the promotion of risk awareness and the definition of key players as well as responsibilities in the course of the emergency preparedness planning process [14].

**Figure 1.** Structuring of the procedure for emergency preparedness planning in water supply based on [12,13,15].

A risk analysis structures and objectifies the collection of information on existing and potential risks to the water supply [14]. The analysis considers the reasons and causes of risks, examines the possible effects and determines the framework within which these consequences can occur [16]. In addition, risk analysis provides the basis for effective and efficient use of limited resources by comparing the various identified risks of processes and components of water supply.

Preventive measures contribute to the reduction of risks for critical processes. They also contribute to achieve operational protection goals and thus raise the barrier for events with crisis potential in the facility [12]. In this way, the number of crisis-prone events can be minimized or the intensity of the events can be reduced.

The processes of crisis managemen<sup>t</sup> help to protect facilities and thus critical infrastructures and the population. Interactions exist with risk management, since not all risks can be reduced by risk-minimizing measures and a residual risk always remains [12]. Crisis managemen<sup>t</sup> therefore offers a structure for coping with crises that cannot be prevented [13,17].

The evaluation refers to all phases, i.e., both the examination of points defined in the preliminary planning, the examination of the topicality of the information on existing risks, the examination of the effectiveness of the implemented preventive measures and the examination of the crisis managemen<sup>t</sup> [14]. It should be repeated regularly.

#### **2. Materials and Methods**

The emergency preparedness planning indicator (EPP) developed in this study is based on a number of main, partial and individual indicators. These indicators cover organisational as well as technical aspects of emergency preparedness planning. The contents of the indicator correspond to the processes and components of effective emergency preparedness planning in water supply. For the development and calculation of the EPP, a multi-stage and iterative process according to [18] was carried out (Figure 2).


**Figure 2.** Procedure for the compilation of a composite indicator based on [18].

#### *2.1. Description of Approach for the Development of Composite Indicators*

The theoretical framework of the EPP is the systematic procedure of risk and crisis managemen<sup>t</sup> according to [13,14], the necessary scope of which is described by five process steps. The EPP therefore consists of the five main indicators of preliminary planning, risk analysis, preventive measures, crisis managemen<sup>t</sup> and evaluation. The Table 1 shows the subdivision of the five main indicators and the 19 sub-indicators in total. Appendix A Tables A1–7 show all individual indicators. In addition to a literature study and existing theoretical models, the individual indicators representative for emergency preparedness planning were developed on the basis of expert and stakeholder knowledge in different workshops, thus applying a stakeholder-oriented methodology. This methodology is usually used in the development of composite indicators (e.g., [19,20]), if, as in the present case, their use is intended as a self-assessment tool for municipalities or authorities.


*Sustainability* 

#### *2.2. Data Acquisition and Selection*

The indicators shown in Table 1 are necessary for the quantitative assessment of the status of implementation of emergency preparedness planning. Thus, data are required that allow both a review of the applicability and significance of the EPP and the determination of the status quo in Germany. However, the required information cannot be determined from publicly available data, as this is utility-specific and relevant to the security of the water utility and its services. A targeted assessment is therefore necessary.

#### 2.2.1. Case Study for Verification of Indicator

To check the applicability and significance of the EPP, all indicators were collected on the basis of a questionnaire for a water supply company as a case study. The case study shows a real water supply utility which supplies a total of 230,000 inhabitants in 120 municipalities and districts. In total, the water supply company delivers about 10 million cubic meters of water annually.

#### 2.2.2. Germany-Representative Dataset

In order to assess the status quo of emergency preparedness planning in Germany, an existing dataset of a nationwide survey on emergency preparedness planning in water supply was analysed. The data set was collected in 2015 by means of a partially standardised questionnaire within the framework of the NoWa I research project with the assistance of the federal level in order to obtain a general, supra-regional overview of the current status of emergency planning in the districts and municipalities. As the responsible bodies, the districts and municipalities had to ask for input from the water supply companies to fill out the data collection form, if the information was not already available. In total, a completed survey questionnaire was returned by 194 districts and 166 municipalities. The data thus consists of 360 individual data sets, which contain data from nationwide distributed municipalities and districts with a population of around 39 million inhabitants.

Each dataset contains information on 37 questions concerning different aspects of emergency preparedness planning and existing water supply systems. To determine the status quo, 21 relevant individual indicators from the data entry form with 37 questions were identified and considered in the EPP. In total, the data sets are assigned to twelve of the 19 sub-indicators (Table 1).

A subsequent data collection or data supplement could not be implemented, since the data collection within the framework of the NoWa I project and an additional collection of a representative data set could not be repeated. Thus, the determination of the status quo does not include all identified indicators.

#### *2.3. Imputation of Missing Data*

Missing data impairs the development and evaluation of composite indicators and can lead to a distortion of the results [18]. In the present study, a case-by-case elimination of data sets is only applied if the data that are absolutely necessary for the situation analysis (e.g., allocation of the data set to the municipality) are not available.

The analysed data of the NoWa I project were collected before the methodology of the emergency preparedness planning indicator was developed. They do not include all indicators relevant to the EPP. Since it was not possible to collect such a data set subsequently, the NoWa I data sets were used to determine the indicator, although they did not include complete indicator data sets (as shown in Table 1). The missing individual indicators are therefore not included in the evaluation.

#### *2.4. Normalisation of Data*

In order to be able to compare the indicators of different municipalities or the individual sub-indicators with each other, a normalisation process is necessary. This is especially true if the data sets differ in their units of measurement [18].

The questions with "yes-no" or "yes-partial-no" possible answers are converted into a [0,1] scale. Likert scales with a given answer scale were also transformed into a [0,1] scale. The answer option "not known" was equated to the answer option "no", since this is equivalent in terms of content for the evaluation of the indicator. This was necessary because the originally planned survey in the NoWa I project had a different assignment of the questions.

#### *2.5. Weigthing and Aggregation*

The individual indicators are integrated into the composite indicator with different weightings (see Equation (1)). This is because the composite indicator is calculated by the weighted sum of its main indicators (see Equation (2)). For the contingency planning indicator, the weighted sum of the five main indicators is determined. As described in [39], there are several ways to determine the weighting of composite indicators.

$$CI = \sum\_{j=1}^{m} x\_j X\_j \tag{1}$$

$$EPP = \mathbf{x}\_{\text{PP}} \cdot \mathbf{PP} + \mathbf{x}\_{\text{RA}} \cdot \mathbf{RA} + \mathbf{x}\_{\text{PM}} \cdot \mathbf{PM} + \mathbf{x}\_{\text{CM}} \cdot \mathbf{CM} + \mathbf{x}\_{\text{E}} \cdot \mathbf{E} \tag{2}$$


For the EPP this paper compares the results of a statistical and an expert-based weighting approach. The main difference between the two approaches is how the indicator weights are derived. Since the weighting of the main and sub-indicators significantly influences the result of the EPP, the composite indicator is determined with identical main and sub-indicators for both weighting approaches.

If the main indicators are equally weighted, they are equally included in the composite indicator. Due to the different weighting of the individual summands (main indicators or sub-indicators), they are assigned a differentiated significance for the composite indicator.

To determine the expert-based weighting, the weights were derived from expert opinions. Using a budget allocation approach, fourteen experts with different specialist backgrounds were asked to assess the main and sub-indicators in a questionnaire using a Likert scale according to their relevance for target-oriented emergency preparedness planning. The weightings of the main indicators derived from the equal distribution and from the expert opinions are presented in Section 3.1.
