1 Introduction

Museum indoor space is a special environment in which both people’s health and collection safety is of concern. The safety of museum collections is achieved by protecting them from accidental issues like fire, earthquake, flood, human error and crime [1,2,3], as well as daily environmental risk factors like extreme temperature (T), relative humidity (RH), illuminance and hazardous materials [4,5,6,7,8,9,10]. The work of a conservation scientist usually includes focusing on the daily environmental conditions and making proper treatments before actual damage occurs to collections, which is called preventive conservation.

There are always different kinds of collections in comprehensive museums. For example, people may find heritage items made from stone, bronze, iron or wood in a historical museum, while they may also see animal bones, plant specimens and fossils in a natural museum. Conservation scientists need to be familiar with physical and chemical properties of different materials so that they can make proper decisions when designing the preventive conservation plan and treating the potential risks. The ideal range of the most important factors in museum indoor environment under the climate of Shanghai is listed in Tables 1 and 2, which are based on the Chinese National Standards GB/T 18883-2002 [11] and GB/T 23863-2009 [12] and partly refer to the ASHRAE Handbook [13].

Table 1 Ideal range of RH and illuminance for conserving different materials
Full size table
Table 2 Ideal range of other key factors for conserving museum collections
Full size table

One significant fact is that objects made from different materials are sometimes placed in the same showcase [14,15,16] due to various reasons, such as they were found in the same place, they are of the same age, or they belong to the same donor. Compared with the ideal case in which only items made from the same material are placed in the same showcase, real cases in comprehensive museums make conservation scientists facing much more complicated situations and they have to be very cautious when making any decisions.

Other factors conservation scientists have to concern include [5, 17, 18]: (i) The outdoor climate; (ii) The structure of the museum construction, especially the exhibition hall; (iii) The performances of air conditioners, humidifiers and humidity controlling agents; (iv) The tightness of showcases; (v) Balance between heritage conservation and display effect, especially concerning illuminance; (vi) Attendance, which is related to the concentration of carbon dioxide in the hall and the indoor T and RH; (vii) The placement of environmental monitors, which has a significant effect on the validity of measured data.

Various kinds of environmental monitoring system have been developed in the recent years for museum use [4, 19,20,21,22,23,24]. These systems are typically based on wireless technique, featuring high flexibility, real-time monitoring and remote data accessing. The commonest monitoring items include T, RH, carbon dioxide, volatile organic compounds, light illuminance, etc.

RH is probably the most important environmental factor influencing the condition of cultural heritage [25, 26]. Water is a good medium for some chemical reactions, playing a role as solvent or catalyst [27, 28], especially promoting the corrosion of metallic objects [25, 29,30,31]. Organics tend to mildew under high RH [15, 25, 32,33,34,35]. On the other hand, water molecules tend to be adsorbed on hydrophilic surfaces, building the microstructure of many organic objects like paper and textile [35, 36], so that an extremely low RH may lead to dehydration and then the deterioration of these objects [25, 35]. Since museums on the world differ significantly in local climate, collection property, building scale and structure, and financial situation, it is impossible to set a single mode or standard for museums to limit RH strictly; instead, recommended standards developed by authoritative institutions are more practicable for specific museums. The control of RH in museums can be achieved by using humidifiers [37,38,39] or passive methods [39]. As the major passive method, various RH controlling agents have been reported, which are suitable for different specific conditions [40,41,42,43,44]. Saturated salt solutions are another passive method which is also effective in treating pollutants like acids and aldehydes [45]. Last but not least, isolating the display space from inside–outside material and energy exchange using physical isolation measures like sealed door or window, or solar control glass combined with light internal drape [46], is a classical but effective passive method.

In this work, the author describes the environment-related work in a special exhibition in Shanghai Museum. One aim of this work is to provide conservation scientists and designers with an example in which various real factors having effects on monitoring and controlling of T and RH are present so that the staff would be more experienced in treating the similar cases. Another aim is to develop a method for assessing and predicting the environmental risk caused by T, RH and their fluctuation, as the risk management is currently of interest to conservation scientists. In the following section of the article, Methods and settings, detailed information about collections, showcases and monitoring and controlling settings in the exhibition, as well as the data analysis idea, are shown. The next section, Results, in which raw data of T and RH for each showcase during the exhibition are provided, is divided into four parts based on the timeline. Factors influencing monitoring and controlling of the micro-environment, including outdoor climate, exhibition hall structure, attendance, showcase structure, showcase tightness, showcase volume, monitor position and object combination modes, are discussed in the Discussion section, and the temperature control results are assessed in the same section. It is considered that although the special exhibition featured long period and complicated showcase setting, the temperature control target is reached in general, while some defects, such as unreasonable showcase structures and settings and the consequent unsatisfactory monitor position and difficulty in RH controlling, also existed but could be avoid. The two RH controlling methods, as passive methods and humidifiers, are compared in the Discussion section, and the former is regarded as having a better effect in inhibiting the RH fluctuation. At the end of the Discussion, a method for assessing and predicting the risk caused by T or RH fluctuation is presented. It is based on analyzing historical and present data and is considered simple and practical for predicting the probability of abnormal fluctuation provided that the basic conditions are constant. The study provides a comprehensive example from which both conservation scientists and exhibition designers could benefit, especially for those not experienced enough for participating in large exhibitions. The study also promotes the understanding of environmental risk assessment and prediction, which is meaningful to researchers of the conservation field.

2 Methods and settings

Various items including stone, jade, bronze, coin, textile, furniture, bamboo product, pottery and porcelain, and traditional Chinese painting and calligraphy were shown in this exhibition. The show time started on March 2022 and ended on January 2023. Air conditioners, humidifiers and three kinds of passive humidity controlling methods were applied to this exhibition.

T and RH of the exhibition hall in Shanghai Museum, China, were set as 20.0 °C in winter (or 22.0 °C in summer) and 55.0%, respectively, using the heating ventilation and air conditioning system (HVAC). Totally 53 showcases were used in this special exhibition. Since showcases 37 and 38 were combined into one, there were actually 52 showcases. Among them, 33 showcases were filled with single kind of objects, while other 19 showcases were applied to placing at least two kinds of objects. MW302GA-HN T-RH monitors produced by Microwise System Co., Ltd., China, were placed into most showcases. The working cycle was set as 10 min, which means for each monitor one set of T and RH data was obtained every 10 min. As showcase 25 was divided into three parts (without isolation) to place different objects, three T-RH monitors were placed into it, while showcase 41 was not equipped with T-RH monitor because the inner item porcelain is not sensitive to T and RH fluctuation. Different RH controlling methods, including dedicated humidifier, desiccant, RH controlling agent (CA), and an experimental adsorbent (EA), were applied to different showcases considering the property of inner items. Twenty-three humidifiers of different powers (produced by Chongqing Zillions Co., Ltd., China) were applied to different showcases considering their volume, while desiccant (produced by Shanghai Bohuan Industrial Co., Ltd.), CA (produced by Shanghai Bohuan Industrial Co., Ltd.) and EA (provided by Tongji University, China) of proper dosages were placed into the corresponding showcases considering their volume as well. EA was tested by using the Oddy Test and it reached the “permanently usable” level prior to its application. The setting of showcases and the corresponding monitoring and controlling measures are listed in Table 3.

Table 3 Settings of showcase, T-RH monitor and RH control method
Full size table

Six types of showcase were used in this exhibition (see Fig. 1): (i) Tall showcase with full-size vertical glass and without any platforms, which was used to show vertical Chinese calligraphy and painting scrolls or furniture; (ii) Showcase with full-size vertical glass and equipped with a hollow desk, which was used to place flat objects like paper; (iii) Showcase with half-size vertical glass and equipped with a full-size inclined platform, which was used to show small flat objects like paper money and coins; (iv) Showcase with half-size vertical glass and equipped with a full-size horizontal platform, which was used to place most types of small objects; (v) Central low showcase with full-size horizontal glass and platform, which was used to show featured small objects; (vi) Specially shaped central showcase 15 with half-size vertical glass and equipped with a full-size horizontal platform, which was used to show two traditional Chinese musical instruments. Showcase types i and ii were equipped with humidifiers hidden behind the artificial wall. Considering the visual effect, designers sealed and covered the access panels for humidifiers using lime and paint. On the other hand, desiccant, CA and EA were typically placed under the platform in showcase types iii, iv and v (some inconspicuous holes were made on the platform to make the two space open to each other), which met the requirement of visual effect.

Fig. 1
figure 1

Typical showcases used in the special exhibition. a Type (i), b Type (ii), c Type (iii), d Type (iv), e Type (v), f Type (vi)

Full size image

Illuminance on different objects was measured and adjusted based on Table 1 in principle after the objects had been in place.

All materials used in the exhibition, including panel, glue, sealing strip, textile, etc., were tested by using the Oddy Test and they reached the “permanently usable” level prior to their application. Concentrations of HCHO and volatile organic chemicals were also qualified during the exhibition. Thus, the content of indoor air contaminants is not discussed in this study.

The whole exhibition period was divided into four stages to learn the variation of T and RH in different time periods. Massive raw monitoring data were given and some series of data were transformed into specific values to conduct various analyses. Factors including outdoor climate, exhibition hall structure, attendance, showcase structure, showcase tightness, showcase volume, monitor position, monitoring availability and object combination modes were discussed in order to assess their effect on monitoring and controlling of T and RH. Furthermore, the corresponding range of maximum fluctuation of temperature in any 24 h (TFmax) (or maximum fluctuation of RH in any 24 h, RHFmax) to each value of average temperature fluctuation calculated during the whole period (TFaverage) (or average RH fluctuation calculated during the whole period, RHFaverage) was determined. The statistic range included all the showcases. The ratio of abnormal showcase to total showcase at each TFaverage (or RHFaverage) value was calculated. The “abnormal” showcase was identified based on TFmax > 2.0 °C, or RHFmax > 5.0%. Each ratio equals to the probability of abnormally high fluctuation under the corresponding TFaverage (or RHFaverage) value.

In order to judge if a factor is significant to the results of monitoring or controlling of T and RH, the One-way Analysis of Variance (ANOVA) was conducted by using the Microcal Origin 6.0 Professional software. The significance level was set as 0.05. The factors tested included showcase position on TFaverage and RHFaverage; attendance on TFaverage, TFmax, RHFaverage and RHFmax; showcase volume on TFaverage; and different controlling methods on RHaverage and RHFaverage at different RH targets. The relevant p-values were given in the corresponding places.

For learning the correlation between outdoor climate and microclimate in showcases, Tmax, Tmin and RHaverage for the showcases in each month were recorded or calculated first. Then, average values of Tmax (Tmax,average) and Tmin (Tmin,average) for all the showcases in the same month were calculated. The correlations between the above values calculated and the corresponding outdoor climate data of Shanghai were tested also by using the Microcal Origin 6.0 Professional software. The coefficients were obtained and given.

3 Results

3.1 From March 10 to April 13

The exhibition started on middle March 2022. However, due to the needs of epidemic prevention and control, the museum was closed soon after that. During the closing period, T and RH monitoring and controlling devices kept the 24 h working mode, and the light sources were switched off unless in the daily inspection time when the security guards examined the collections, while other non-front-line staff members including the conservation scientists could only see the monitoring data online but they could not come to the field except for serious and urgent cases. On middle April the storage department of the museum asked for a survey of the environmental data in the past month, thus the data were provided, as shown in Tables S1 and S2. Some information was obtained from the above two tables:

  1. (i)

    Five monitoring devices did not work from the beginning of the exhibition.

  2. (ii)

    In most showcases (43 of 46 showcases equipped with working monitors), the average temperature was located at the range 20.0 °C ± 2.0 °C. Monitors in showcases 25A, 25B and 29 detected that the average temperature values reached 22.2, 22.1 and 22.5 °C, respectively.

  3. (iii)

    In all the showcases equipped with working monitors, the 35-day average value of the temperature fluctuation in each 24 h (TFaverage) did not exceed 2.0 °C. If the maximum temperature fluctuation in one day during the 35-day period (TFmax) was applied to assessing the temperature fluctuation, seven showcases had the case that the TFmax value exceeded 2.0 °C; showcases 13 and 33 were found to have the most abnormal days of TFmax value, as 7 and 5 days, respectively.

  4. (iv)

    Monitors in showcases 3, 4, 5, 7, 11, 14, 17, 29, 40, 47, 48, 50, 51, 52 and 53 detected that the average RH values exceeded the upper limit of the ideal range. The average RH values in showcases 3 and 29 were far from the ideal range, while the exceeded volume in showcase 5 was 8.1%, and the exceeded volumes in the other showcases were not higher than 5.0%.

  5. (v)

    In most showcases equipped with working monitors, the 35-day average value of the RH fluctuation in each 24 h (RHFaverage) did not exceed 5.0%. The RHFaverage values in showcases 3, 11 and 27 exceeded 5%, as 5.2%, 6.5% and 5.5%, respectively. If the maximum RH fluctuation in one day during the 35-day period (RHFmax) was applied to assessing the RH fluctuation, 11 showcases had the case that the RHFmax value exceeded 5.0%; showcases 3, 11 and 27 were found to have the most abnormal days of RHFmax value, as 19, 25 and 17 days, respectively.

3.2 From April 14 to August 31

During this period, two events are noticeable: the museum was reopened to public from July 1; and the humidifier for showcase 3 and the T-RH monitor in showcase 29 were replaced by new ones on August 29 during the treatment time. The data collected during this period are shown in Tables S3 and S4. Some information was obtained during this period:

  1. (i)

    Monitors in showcases 5, 6, 13 and 36 stopped working on August 29, August 26, August 29 and May 4, respectively, due to low battery.

  2. (ii)

    In all the showcases equipped with working monitors, the average temperature was located at the range 22.0 °C ± 2.0 °C, which is ideal for the exhibition in summer.

  3. (iii)

    In all the showcases equipped with working monitors, the TFaverage value did not exceed 2.0 °C. If the TFmax value was applied to assessing the temperature fluctuation, 21 showcases had the case that the TFmax value exceeded 2.0 °C; showcase 7 had the most abnormal days of TFmax value (10 days).

  4. (iv)

    Monitors in showcases 3 (before treatment on August 29), 4, 5, 7, 11, 14, 17, 24, 25A, 25B, 25C, 29 (before treatment on August 29), 40, 47, 48, 50, 51, 52 and 53 detected that the average RH values exceeded the upper limit of the ideal range. The average RH value in showcase 29 were far from the ideal range, while the exceeded volumes in showcases 3, 11 and 47 were 6.7%, 5.6% and 8.3%, respectively, and the exceeded volumes in the other showcases were not higher than 5.0%. Monitor in showcase 33 detected that the average RH value was lower than the lower limit of the ideal range by 5.3%, while monitors in showcases 6, 9, 15 and 30 detected that the average RH values were also lower than the lower limit of the ideal range, but the lowered volumes were not higher than 5.0%.

  5. (v)

    After treatment on August 29, both showcases 3 and 29 showed a qualified condition on the average RH index.

  6. (vi)

    In most showcases equipped with working monitors, the RHFaverage value did not exceed 5.0%. The RHFaverage value in showcase 11 exceeded 5.0% (6.7%). If the RHFmax value was applied to assessing the RH fluctuation, 20 showcases had the case that the RHFmax value exceeded 5.0% before treatment on August 29; showcases 3, 8, 11, 17, 27 and 29 were found to have the most abnormal days of RHFmax value, as 43, 23, 119, 30, 32 and 25 days, respectively. After treatment on August 29, both showcases 3 and 29 showed a qualified condition on the RHFmax index.

3.3 From September 1 to September 30

During this period, one noticeable event is that the humidifier for showcase 11 was checked and debugged on September 26. The data collected during this period are shown in Tables S5 and S6. Some information was obtained during this period:

  1. (i)

    Monitors in showcases 2, 12, 19 and 46 stopped working on September 18, September 27, September 28 and September 13, respectively, due to low battery. Monitor in showcase 29 stopped working on September 5 due to wireless connection lost.

  2. (ii)

    In all the showcases equipped with working monitors, the average temperature was located at the range 22.0 °C ± 2.0 °C, which is ideal for the exhibition in summer.

  3. (iii)

    In all the showcases equipped with working monitors, the TFaverage value did not exceed 2.0 °C. If the TFmax value was applied to assessing the temperature fluctuation, 16 showcases had the case that the TFmax value exceeded 2.0 °C; showcase 25 A had the most abnormal days of TFmax value (4 days).

  4. (iv)

    Monitors in showcases 4, 7, 11 (before and after treatment on September 26), 14, 17, 24, 25A, 25B, 25C, 29, 40, 47, 48, 50, 51, 52 and 53 detected that the average RH values exceeded the upper limit of the ideal range. The exceeded volumes in showcases 11 (before treatment on September 26) and 47 were 5.7% and 8.2%, respectively, and the exceeded volumes in the other showcases were not higher than 5.0%. Monitor in showcase 33 detected that the average RH value was lower than the lower limit of the ideal range by 6.6%, while monitors in showcases 9, 15 and 30 detected that the average RH values were also lower than the lower limit of the ideal range, but the lowered volumes were not higher than 5.0%.

  5. (v)

    After treatment on September 26, the average RH of showcase 11 decreased by 5.5%, but it was still slightly higher than the upper limit of the ideal range.

  6. (vi)

    In most showcases equipped with working monitors, the RHFaverage value did not exceed 5.0%. The RHFaverage value in showcase 11 exceeded 5.0% (6.1%, before treatment on September 26). If the RHFmax value was applied to assessing the RH fluctuation, 20 showcases had the case that the RHFmax value exceeded 5.0% before treatment on September 26; showcase 11 was found to have the most abnormal days of RHFmax value, as 19 days. After treatment on September 26, showcase 11 showed a qualified condition on the RHFaverage index, while the RHFmax index decreased by 3.1% but it was still unqualified.

3.4 From October 1 to January 8

The data collected during this period are shown in Tables S7 and S8. Some information was obtained during this period:

  1. (i)

    Monitors in showcases 1, 4, 7, 8, 14, 28, 35 and 52 stopped working on October 25, October 1, November 6, November 15, December 2, October 15, October 21 and October 8, respectively, due to low battery.

  2. (ii)

    In most showcases equipped with working monitors, the average temperature was located at the range 20.0 °C ± 2.0 °C, which is ideal for the exhibition in winter. Monitors in showcases 4, 8, 10, 25A, 25B, 28, 35 and 49 detected that the average temperature values reached 22.7, 22.2, 22.1, 22.4, 22.3, 22.4, 22.7 and 22.2 °C, respectively.

  3. (iii)

    In all the showcases equipped with working monitors, the TFaverage value did not exceed 2.0 °C. If the TFmax value was applied to assessing the temperature fluctuation, five showcases had the case that the TFmax value exceeded 2.0 °C; showcases 25A and 35 had the most abnormal days of TFmax value (both were 4 days).

  4. (iv)

    Monitors in showcases 4, 7, 14, 17, 24, 25A, 25B, 25C, 40, 47, 48, 50, 51, 52 and 53 detected that the average RH values exceeded the upper limit of the ideal range. The exceeded volume in showcase 47 was 5.7%, and the exceeded volumes in the other showcases were not higher than 5.0%. Monitor in showcase 9 detected that the average RH value was lower than the lower limit of the ideal range, but the lowered volume was not higher than 5.0%.

  5. (v)

    In all the showcases equipped with working monitors, the RHFaverage value did not exceed 5.0%. If the RHFmax value was applied to assessing the RH fluctuation, 14 showcases had the case that the RHFmax value exceeded 5.0%; showcases 11, 17 and 27 were found to have the most abnormal days of RHFmax value, as 30, 16 and 22 days, respectively.

4 Discussion

4.1 Outdoor climate

The outdoor climate is probably the first and the most important consideration when designing the indoor climate-control system of the museum building, and it has a permanent influence on the indoor environment [47,48,49,50]. The far-reaching influence of global climate change should also be considered [51]. The outdoor temperature information of Shanghai is listed in Table 4 according to the daily weather reports downloaded at http://www.moji.com [52]. It can be learnt from the data that the hottest time during the exhibition was in July and August, while the coldest time was in December and early January. The distinctive climate not only shows an obvious annual temperature fluctuation, but also indicates a high RH fluctuation between different months (see Table 5). According to the coefficients given in Table 6, it can be reached that the Tmin,average value for the microclimate has a highly positive correlation to the outdoor Tmin value, and the Tmax,average value for the microclimate has a moderately positive correlation to the outdoor Tmax value, while the RHaverage value for the microclimate has a moderately negative correlation to the outdoor RHaverage value, indicating the effect of outdoor climate on the micro-environment.

Table 4 Outdoor temperature values (°C) during the exhibition
Full size table
Table 5 Average RH (%) of Shanghai during 1980–2010 according to the online database at http://m.weather.com.cn [53]
Full size table
Table 6 Correlation between the outdoor climate and the microclimate 
Full size table

4.2 Exhibition hall structure

Some studies have pointed out that the structure of exhibition hall has a potential effect on the control of T and RH [5, 17]. As shown in Fig. 2, there was a preface hall when people accessed into the exhibition. External T fluctuation and RH fluctuation might have an obvious effect on the environment of the preface hall and the surrounding showcases, especially showcases 1, 2, 3, 4, 8, 47, 52 and 53 (Zone A). In contrast, showcases 29–42 (Zone B) were located at the deepest section of the whole hall, which might be favorable for keeping a low fluctuation. In fact, TFaverage for Zone B during the whole exhibition period (0.7 °C) was significantly lower than that for Zone A (0.9 °C) (p = 0.022), indicating that the showcase position plays an important role in reducing the effect of outdoor TF. However, although RHFaverage for Zone B during the whole exhibition period (1.8%) was also lower than that for Zone A (2.4%), the difference was insignificant (p = 0.11), indicating that the controlling measures were relatively more successful in offsetting the effect of the exhibition hall structure on RHFaverage.

Fig. 2
figure 2

The exhibition hall structure

Full size image

4.3 Attendance

Since the special exhibition was very popular in Shanghai, the attendance number often reached the upper limitation. According to the references [5, 15, 54], visitors may lead to an increasing of RH, especially in summer when the exhibition room is full of people. Since the museum was closed from March 10 to June 30, the attendance during this period was nearly zero, so the T and RH data during this period can be regarded as a background value without the effect of attendance. As July 1 was the open day, T and RH data from June 24 to 30 were used to characterize the condition before opening, while data from July 1 to 7 were used to characterize the condition after opening. The two consecutive short periods ensure the similarity of weather and the validity of data. Some information can be found from Table S9: (i) After opening, the temperature in every showcase had a slight increase; (ii) After opening, the RH in most showcases had a slight decrease or kept unchanged, while in showcases 9, 10, 14, 15, 16, 17, 19, 24, 25A, 25C, 26, 38, 40, 48, 49, 50, 52 and 53 it presented a slight increase; (iii) After opining, the temperature fluctuation in most showcases had an increase or remained unchanged, while in showcase 2 it had a decrease; (iv) After opening, the RH fluctuation in most showcases had an increase or remained unchanged, while in showcases 2, 11, 39 and 53 it presented a decrease. It can be inferred from the above results that: (i) Attendance plays a role in increasing the temperature in showcases; (ii) The variation of RH caused by attendance can occur in any showcases regardless of the RH controlling methods; (iii) Attendance has a significant effect on increasing both TFaverage (p = 3.81 × 10−9) and TFmax (p = 4.42 × 10−6) in showcases; (iv) Attendance also has a significant effect on increasing RHFmax (p = 0.015) in showcases, but does not have a significant effect on increasing RHFaverage (p = 0.12) in showcases. However, the effect of attendance in winter, and the reason for abnormal cases occurring in showcases 2, 11, 39 and 53, need further studies. Furthermore, according to the literature [55], the movement of visitors not only causes heat and moisture transfer, but also leads to contaminant diffusion, which is of referential value for the future work.

4.4 Showcase structure

Passive RH controlling methods are usually not suitable for showcase types i and ii, because these showcases are usually large so that the dosage of RH controlling agents must also be large [14], and the unshielded space makes it difficult to select a suitable position for the agents. On the other hand, sealing the manholes made it actually impossible to access the humidifiers without destroying the artificial wall body. Unluckily, as the measured data indicated, showcases 3, 11 and 29 really had trouble in RH controlling, which might be directly associated with the working condition of the humidifiers, but staff members had no way to check the machines immediately but wait for an appointed date to open the panel. Moreover, since the special exhibition lasted for a rather long period including four seasons, desiccant, CA or EA used in showcase types iii, iv and v might be expired during the exhibition, while staff members could not replace them facing the strictly closed showcases. In a word, designers have to think over the structure of showcases applied to long-period special exhibitions. Both the access panel for humidifiers and the position for passive RH controlling agents must be easier to access, so that the staff can treat the situation in time.

4.5 Showcase tightness

Showcases can be produced by machine or made manually. The former is preferred from the view of preventive conservation because it favors the formation of a high tightness. Although a tight showcase does not mean an extremely separated space, it provides the inner objects with a considerable protection from the outside air [56, 57], especially the RH fluctuation [18, 58]. However, since most showcases for the special exhibition were made manually using boards rather than integratedly molded, the tightness of them did not always reach the air-tight level of museum showcase according to the Chinese National Standard GB/T 36110-2018 [59]. Thus, the theoretical dosage of desiccant, CA or EA, or the theoretical power of humidifiers, was probably insufficient for controlling the RH in these showcases [14, 60], especially when the showcase space was large.

4.6 Showcase volume

It can be seen from Table 3 that seven showcases were larger than 4 m3. It is not quite suitable to control the RH in these big showcases using passive methods (desiccant, CA or EA), as it may need a quite large volume of them, and they need to be fully distributed to achieve the best effect. Nevertheless, showcase 25 (4.8 m3) was equipped with desiccant because of its structure as a central showcase in one section of the exhibition. Meanwhile, the large space is also a big challenge to the performance of humidifiers, even if some humidifiers are claimed to have the power for controlling a 6 m3 space, because the rather long pipeline set in the showcase makes the airflow exchange difficult. It can be learnt from the above results that showcase 29 was seriously unqualified in RH from March 10 to August 29, which is probably associated with its large volume. According to the research by Yang et al. [61], in a very large showcase (7.5 m length × 0.95 m width × 3 m height) equipped with humidifiers, the upper space is higher than the lower space in temperature, while on the contrary in RH; the central position of the upper space has both the highest temperature and the lowest RH, while the two corners of the lower space have both the highest RH and the lowest temperature. Based on this rule, temperature values measured in many big showcases in the present study, especially in showcase type i, were probably lower than the real values, while RH values measured in these showcases were probably higher than the real values, as the monitors were placed on the ground platform.

Based on the data listed in Table S10, it can be learnt that during the whole exhibition period the average daily temperature fluctuation in showcases > 4 m3 was 0.80 °C, while in showcases ≤ 1 m3 was 0.95 °C, and in the rest showcases (1 m3 < volume ≤ 4 m3) was 0.79 °C. According to the research by Lei [15], the showcase volume is related to the airflow behavior. In brief, air in small central showcases does not diffuse or flow so easily like in larger showcases, which leads to a result that T and RH in small showcases vary more significantly than in larger showcases when the external climate changes. However, the difference of daily TFaverage between small (≤ 1 m3) and middle-size (1 m3 < volume ≤ 4 m3) showcases is significant (p = 0.043), while both the difference between large (> 4 m3) and middle showcases (p = 0.93), and the difference between large and small showcases (p = 0.23), are insignificant. As the middle-size showcases were very similar to the larger ones in TFaverage, they actually can be combined. Thus, in this study 1 m3 can be regarded as the critical volume differentiating showcases of lower or higher temperature fluctuation characteristic. In fact, other variables other than showcase volume, such as light, monitor position and showcase structure, are all responsible for the temperature fluctuation. It is hard to evaluate the effect originated from each single factor in a precise way at the current stage, but it may be achieved after acquiring more data based on more precise experiment designs. On the other hand, since RH in small and big showcases in the present exhibition was controlled through different strategies, the impact of showcase volume on RH fluctuation in these two modes cannot be compared directly. In fact, since the passive methods used in the exhibition had a better effect on inhibiting the RH fluctuation (see Sect. 4.10 below), the impact of showcase volume on RH fluctuation can be offset.

4.7 Monitor position

The placement of monitors directly influences the effectiveness of data measured [47, 62]. Monitors should be placed close to the objects in principle so that the data collected reflect the real condition of the micro-environment according to European Standards EN 15758:2010 [63] and EN 16242:2012 [64] and Frasca et al. [62]. However, visitors will see the monitors in that case, which is undesired on the visiting experience [24, 62]. Considering that, the exhibition designers requested that the monitors be placed at inconspicuous positions, such as corners of a display platform, the position behind a tall object, the position under a display platform, or corners on the ground platform. It was more difficult for people to see the monitors (although they could find them in many cases if they carefully examined the showcases), but at the same time the effectiveness of the collected data decreased. In the tall showcases singly displaying Chinese calligraphy and painting, the monitors were placed on the ground platform actually far from the body of the collections, which made the collected RH data deviate from the real values, and was probably one reason for the abnormally high RH detected in showcase 29 from March 10 to August 29. On the other hand, since the Chinese calligraphy and painting collections were displayed vertically, it would be perfect to have three monitors placed on different heights for a collection to learn the real RH condition comprehensively. However, it can hardly be done due to visiting experience and cost considerations. Another regret is that there was not a suitable position in the exhibition hall for placing a T-RH-CO2 three-in-one monitor to record T and RH of the exhibition hall as well as concentration of CO2 reflecting the number of attendance. To improve, conservation scientists need to think over a position in the exhibition hall which is about as high as an adult, free of covers but inconspicuous, and close to the core section and far from the air outlets. It is probably hard to find a position meeting the requirements perfectly when the exhibition hall is completed, so conservation scientists may need to communicate with designers as early as possible to reserve or set such a position in the designing stage.

4.8 Object combination modes

As shown in Table 3, 19 of the 52 showcases were filled with at least two kinds of objects. Especially, 9 of the 19 showcases were filled with rather different collections, including showcases 10 (bronze, jade, sealing clay, bamboo), 12 (Chinese calligraphy, animal bone, bronze), 19 (jade, Chinese calligraphy, bronze), 25 (bronze, sealing clay, jade, animal bone), 27 (jade, bronze, stone, Chinese calligraphy), 35 (ivory, iron, Chinese painting, stone), 40 (bronze, jade, turquoise), 46 (paper money, silver coin) and 50 (bronze, Chinese painting). This organic-inorganic mixed placing mode is a rather difficult situation for the conservation scientists, as it is impossible to meet the RH requirements of both materials in the same showcase, especially when the upper limits of the ideal RH ranges differ by 15% (metal and organics). As the combination modes were based on the donors, it was impossible to rearrange the objects to make each showcase filled with only one kind of objects, but to focus on the most vulnerable kind in each showcase. The organic collections were considered prior to the inorganic objects in this exhibition, but in showcases 25 and 40 the quantity of bronzes was considered first. Obviously, this method does not always arrive at a balance but leads to potential risks to the relatively “sturdy” objects. For the target of preventive conservation, it would be better to make isolated spaces using glass for different collections [14, 58]. However, from the view of an exhibition designer, this solution does harm to the visual effect. Thus, conservation scientists and designers need to communicate continuously with each other to find a better balance. The four points listed below are important: (i) in many cases the situation is unavoidable, the conservation scientists should communicate with the designers as early as possible to learn the basic information such as ratio of different object types, showcase structure and volume; (ii) select the controlling methods based on the above information, the principle is that more vulnerable objects are prior to the others, organics are usually prior to the inorganics, and the ratio should also be considered; (iii) think over and set the position of RH controlling agents or inlet/outlet of humidifiers reasonably, as sometimes the humidity gradient originated from specific position set may be favorable to objects of different types and positions; (iv) observe and record the condition of collections diligently, and be ready to treat the potential events.

4.9 Temperature control

Temperature in the exhibition hall was controlled by the HVAC system. Considering that the weather in Shanghai is very hot in summer and rather cold in winter, engineers set the temperature target as 22.0 °C in summer and 20.0 °C in winter, which is not only comfortable for the visitors as they would not feel very significant differences between outdoor and indoor temperatures, but also safe to the collections as smaller differences between outdoor and indoor temperatures indicate smaller differences in RH and milder outdoor–indoor RH convection. Temperature and temperature fluctuation in showcases were determined by outdoor environment, air conditioner performance and showcase tightness. As shown in Table 7, the temperature values obtained from April 14 to September 30 were slightly higher than those obtained in other time periods, which is associated with both outdoor temperature and indoor temperature setting. The average values were located at the ideal ranges according to Table 2, reaching the temperature control target in general. On the other hand, since both the average values of temperature fluctuation and the average values of maximum temperature fluctuation during different time periods were not higher than 2.0 °C (see Table 8), the target of temperature fluctuation control was achieved as well. According to the ASHRAE Handbook [13], the temperature fluctuation control result could be classified as the AA level. Nevertheless, some showcases did not always reach the target of temperature or temperature fluctuation control, as shown by the data measured, which may be due to bad monitor position, tightness failure or accidental error.

Table 7 Maximum, average and minimum values of temperature (°C) in showcases during different time periods and the average values of maximum and minimum temperature (°C) in different showcases
Full size table
Table 8 Maximum and average values of temperature fluctuation (°C) in showcases during different time periods and the average values of maximum temperature fluctuation (°C) in different showcases
Full size table

4.10 Comparison between passive methods and humidifier

Active and passive methods have different characteristics and are suitable for different cases. The active method, or the humidifier, is suitable for large showcases with full-size glass and no solid platforms provided that there is continuous supply of electric power [14]. The passive method, or the generalized RH controlling agents, is suitable for small showcases with half-size glass and solid platforms [23], especially for those showcases without continuous supply of electric. Both methods need an inconspicuous space hidden in the showcase. Also, both methods are only effective in a limited volume of showcase. The setting of the methods had been thought over for using in the exhibition. As shown in Table 9, the following results were obtained: (i) When the RH target was set as ≤ 50.0%, both desiccant and humidifier could hardly make the real RH meet the requirement, meanwhile the effect of humidifier was better in general; (ii) When the RH target was set as 55.0% ± 5.0%, CA, EA and humidifier could make the real RH qualified in general, meanwhile the real RH controlled by humidifier, EA and CA was degressive in order; (iii) When the RH target was set as 60.0% ± 5.0%, the real RH controlled by humidifier was significantly higher than that by CA (p = 1.13 × 10−4), meanwhile the effect of humidifier was better; (iv) The measured RH values turned higher and higher along with time; (v) The real RH controlled by CA, EA and humidifier presented a continuous decreasing in the first three periods and a rebound in the last period; (vi) CA had a significantly better performance in controlling the RH fluctuation when the RH target was set as 60.0% ± 5.0%, compared with that as 55.0% ± 5.0% (p = 5.13 × 10−4); (vii) Desiccant had a better performance in controlling the RH fluctuation compared with humidifier did; (viii) CA had a significantly better performance in controlling the RH fluctuation when the RH target was set as 60.0% ± 5.0%, compared with humidifier did (p = 6.37 × 10−8); (ix) The RH fluctuation controlled by humidifier, EA and CA was degressive in order. It can be inferred from the above results that: (i) The desiccant volume experienced a gradual exhaustion along with time, which is in accordance with the reference [65]; (ii) The measured RH is associated with the outdoor climate and the artificial control; (iii) The passive methods are better than the humidifier in inhibiting the RH fluctuation; (iv) The passive methods have a better performance when the RH target is increased from ≤ 50.0% to (60.0 ± 5.0)%; (v) CA is better than EA in inhibiting the RH fluctuation regardless of the significant differences between their dosages, but EA may have a better performance if the dosage is higher.

Table 9 Average RH values (%) achieved by using different controlling methods
Full size table

4.11 Risk assessment and prediction

As a kind of research originated from practice and raised to theoretical level, risk assessment and prediction of museum environment has received much attention and several studies have been reported [66,67,68,69,70,71]. Similar to the above studies, one key target of analyzing the data collected in this exhibition was to develop a simple method or find a practicable index for assessing the current environmental risk in the museum exhibition and predicting the future risk. Some valuable indexes have been introduced in studies [46, 72]. When analyzing the daily T and RH data, the author was interested in the relation between the average temperature fluctuation of a showcase during a selected period (TFaverage) and the maximum fluctuation of the showcase during this period (TFmax). It is easy to see that: (i) A higher TFaverage usually led to a higher TFmax; (ii) TFaverage was close to half of TFmax in many cases, but exceptions or higher deviations also existed. The relation between TFaverage and TFmax range was obtained when the above thinking was extended to all the showcases in the whole exhibition period, as listed in Table 10. It can be seen that: (i) TFaverage ranged from 0.3 to 1.5 °C in the whole exhibition period; (ii) TFmax range turned higher along with the increase of TFaverage in general; (iii) Most TFaverage values ranged from 0.6 to 0.9 °C; (iv) The critical value at 1.0 °C of TFaverage was very clear when calculating the ratio of abnormal showcase to total showcase if the critical value of TFmax was set as > 2.0 °C. As 2.0 °C is the upper limit of temperature fluctuation in 24 h, the above statistics can be applied to calculating the probability of abnormally high temperature fluctuation in any showcases in the whole exhibition period. On the other hand, when singly considering the temperature fluctuation factor, the showcase environment can be regarded as low risk level when TFmax ≤ 2.0 °C, medium risk level when 2.0 °C < TFmax ≤ x °C, and high risk level when TFmax > x °C. Based on the above method, it is obtained that when TFaverage was controlled as ≤ 1.0 °C, the probability of medium and high risk was 16.3%; when TFaverage was > 1.0 °C, the probability of medium and high risk significantly increased to 82.8%. The conclusion can be applied to predicting the future risk in the exhibition if it continues and the environmental conditions remain unchanged.

Table 10 The average value of temperature fluctuation in the whole exhibition period and the corresponding range of maximum temperature fluctuation in any 24 h
Full size table

Similarly, the above method can be applied to assessing and predicting the risk caused by RH fluctuation. It can be seen from Table 11 that: (i) RHFaverage ranged from 0.3 to 6.7% in the whole exhibition period; (ii) RHFmax range turned higher along with the increase of RHFaverage in general; (iii) Most RHFaverage values ranged from 1.8 to 2.4%; (iv) The critical value at 2.5% of RHFaverage was not quite clear when calculating the ratio of abnormal showcase to total showcase if the critical value of RHFmax was set as > 5.0%. As 5.0% is the upper limit of RH fluctuation in 24 h, the above statistics can be applied to calculating the probability of abnormally high RH fluctuation in any showcases in the whole exhibition period. On the other hand, when singly considering the RH fluctuation factor, the showcase environment can be regarded as low risk level when RHFmax ≤ 5.0%, medium risk level when 5.0% < RHFmax ≤ x%, and high risk level when RHFmax > x%. Based on the above method, it is obtained that when RHFaverage was controlled as ≤ 2.5%, the probability of medium and high risk was 20.1%; when RHFaverage was > 2.5%, the probability of medium and high risk significantly increased to 88.4%. The conclusion can also be applied to predicting the future risk in the exhibition if it continues and the environmental conditions remain unchanged.

Table 11 The average value of RH fluctuation in the whole exhibition period and the corresponding range of maximum RH fluctuation in any 24 h
Full size table

Furthermore, the probability of medium and high risk in different time periods was calculated and listed in Table 12. It can be seen that in the second stage both the probability of medium and high temperature fluctuation risk and the probability of medium and high RH fluctuation risk increased significantly compared with those in the first stage, indicating that time is an important factor in the environmental risk assessment and prediction system. However, in the last stage the risk values did not further increase compared with those in the third stage, which indicates that the environmental risk caused by T and RH fluctuation is relevant to the season change and the corresponding artificial control, while the risk values may come to solider values if the exhibition continues.

Table 12 Ratio of abnormal showcase to total showcase in different time of the exhibition
Full size table

Based on the above discussion, a simple method for predicting the risk originated from temperature fluctuation or RH fluctuation is established. First, conservation scientists need to collect enough valid T and RH data in a long period continuously. The key conditions include placing enough monitors in different showcases at proper position (monitors and showcases are as much as possible), making the monitors working continuously for a long period (the time is as long as possible), and keeping the basic conditions constant (such as lighting, setting of air conditioner and humidifier, etc.). Second, divide the whole period into several parts based on season change or important event points, which is beneficial to obtain more data under different external climates. Third, conduct a series of data statistics and calculation as described in Sect. 2. Each ratio obtained equals to the probability of abnormally high fluctuation under the corresponding TFaverage (or RHFaverage) value. As the basic conditions are kept constant, the calculated probability not only reflects the previous and the current risk, but also can be applied to predicting the future risk.

Last but not least, the above risk prediction method is an empirical method, so that the critical value of TFaverage (or RHFaverage), the TFmax (or RHFmax) range and the probability of risk can be variable along with time, and the conclusions are only valid under the specific environmental conditions in the study. Environmental conditions in more exhibitions will be studied in order to obtain more real data to update the current database, revise the prediction model and extend the method to more applications.

5 Conclusions

This work aims to give conservation scientists and designers an example with various real factors having effects on monitoring and controlling of T and RH, as well as develop a method for assessing and predicting the risk level associated with the value of T, RH and their fluctuation. Although the special exhibition featured long period and complicated showcase setting, the control targets of temperature and temperature fluctuation are reached in general, while some defects, such as unreasonable showcase structures and settings and the consequent unsatisfactory monitor position and difficulty in RH controlling, also existed. The distinctive climate of Shanghai has an important effect on the exhibition hall environment. The inner zone of the exhibition hall is favorable for reducing the effect of outdoor TF, while the controlling measures are relatively more successful in offsetting the effect of the exhibition hall structure on RHFaverage. Attendance plays a role in increasing T, TF and RHF in showcases. TF in small showcases (≤ 1 m3) is higher than in larger showcases (> 1 m3) regardless of the HVAC system, but the difference does not reach a significant level. The passive methods are better than the humidifier in inhibiting the RH fluctuation, and have a better performance when the RH target is increased from ≤ 50.0% to (60.0 ± 5.0)%. CA is better than EA in inhibiting the RH fluctuation if the significant differences between their dosages are not considered, but EA may have a better performance if the dosage is higher. TFaverage and RHFaverage are practicable indexes for assessing the current environmental risk caused by TF and RHF and predicting the future risk. Time is an important factor in the environmental risk assessment and prediction system. In general, the study provides a comprehensive case from which both conservation scientists and exhibition designers could benefit, especially for those not experienced enough for participating in large exhibitions. The study also promotes the understanding of environmental risk assessment and prediction, which is meaningful to researchers of the conservation field. As a beginning of a long-term study project aiming to give conservation scientists a clear and detailed guideline for intervening the museum microclimate based on multiple environmental parameters, this work does not involve deep studies on body conditions of the collections and well-developed mathematical models. Luckily, Shanghai Museum and several other institutions in China are conducting the project and will continue testing the obtained models and revising them based on real body conditions of the collections. Furthermore, it is sometimes difficult to classify the critical or non-extreme values, which is one of the key problems to solve in the project and the solutions may be given along with continuous increasing of the data and updates of the assessment/prediction models.