The National Archives and Records Administration (NARA 2013) changed relative humidity control from a steady 45% rh to a permissible seasonal swing between 30% and 50% rh. They estimated a savings of $650,000 per year in utility costs as well as an increase of 20% in collection lifetime. This section enables quantitative answers to very common questions about sustainable variations on fixed standards: how much do benefits change with adjustments in temperature and relative humidity? How much does risk climb during high temperature and high humidity? Are seasonal adjustments possible for energy saving? What are the risks during retrieval? This section is also a reminder that not only archives, but all collections with low-stability objects listed in Table 5, can benefit from low temperature.

Lower temperature reduces the rate of all forms of chemical decay. For the most rapidly decaying organic materials (right-hand columns of Table 5), the dominant mechanism is acid hydrolysis, which increases strongly with the acidity of the material, and increases with moisture content (Zou et al. 1996). Moisture content, in turn, depends on relative humidity. The consensus is that the rate of decay (or its reciprocal, lifetime) is a product of an acidity factor, a temperature factor, and a relative humidity factor (which sometimes includes a correction dependent on temperature).

(1)

where L is lifetime, in years.

For our purposes, acidity is a given, and only temperature and relative humidity can be controlled. The temperature function is an Arrhenius equation:

(2)

where

Michalski (2002) compiled data from reviews of activation energies (notably reviews by Nishimura [1996]) for paper, film, and photographic dyes, as well as further individual studies of magnetic media and the yellowing of varnish. More than three-quarters of all the studies of paper degradation, acetate film degradation, and dark fading of dyes fit within an E_a range of (2.7 to 4) × 10⁷ ft · lb_f /lb mole. In Figure 5, this range is shown by the shaded area. Michalski (2002) further showed that this range of E_a can be derived with no consideration of a specific material, but simply by examining the kinetics of a chemical process that requires several decades to proceed at room temperatures. Thus, both data and theory suggest that this E_a range can be used to estimate the benefits of cold storage, and the risks from high temperature, for all organic materials suspected of being low or very low stability (Table 5). Michalski (2000, 2002) selected a middle value for mixed collections at 3.35 × 10⁷ ft · lb_f /lb mole, shown by the heavy black line in Figure 5. For decay of polyester polyurethane (the weak link in magnetic media, and a popular material with artists in the late twentieth century) and for yellowing of natural resins, the activation energies fall slightly lower, (2 to 2.7) × 10⁷ ft · lb_f /lb mole (between the shaded area and the dashed line in Figure 5.)

Figure 5. Effect of Temperature on Lifetime for Various E_a

Data on the influence of relative humidity are much less extensive than those for temperature, and insufficient to select between differing models. Some authors assume that the true variable is moisture content (Strlic et al. 2015; Zou et al. 1996) and use a complex function of relative humidity that includes a temperature correction.

Three advisory tools are currently available: the preservation index (Reilly 1995), Michalski (2000), and Strlic et al. (2015). These tools can be used to derive equations for a lifetime (L_r) relative to the lifetime at room conditions 68°F, 50% rh. These equations have been used to plot lines of constant relative lifetime (also called isoperms) on the psychrometric chart in Figure 6. The following equations are all arranged with the temperature and relative humidity components separated for clarity, and with R separated so that E_a becomes explicit as the numerator above RT.

Between 20 and 60% rh, differences in the three models are negligible for practical purposes (Figure 6). Beyond this humidity range, the models diverge because of the different functions selected (exponential versus power law), but this is largely irrelevant because relative humidity extremes are usually avoided (mold risk at very high values, and mechanical risks at very low values). The small differences between models on the effect of very low temperature are not because of any differences in opinion about the function (Arrhenius), but on the value of E_a selected, which depends on the particular data set that each author emphasized: acetate film, mixed collections, or acidic paper. The E_as of all three models fit within the shaded area of Figure 5. Essentially, in the range of 20 to 60% rh and for high and low temperatures, any of the three models can be used and will provide the same practical answers.

Figure 6. Lines of Constant Lifetime (Isoperms) for Three Models

A common technical question beyond the scope of fixed standards is the impact on lifetime of out-of-spec events, or seasonal fluctuations. Estimates can assume a simple linear dependence on relative humidity: for instance, if half the year is at 30% rh and half at 60% rh, then the effective annual relative humidity is the average: 45% rh. Temperature dependence, however, is far from linear, and averages cannot be used. The derived equations are general and users can select a preferred E_a, but the graphs and worked examples assume the middle E_a value of 3.35 × 10⁻⁷f t ·lb_f /lb mole (the small differences of E_a of the three models do not make practical differences).

A common query concerns the effect of short periods at higher temperature. The reciprocal of Equation (2) can be used to find the average rate of decay of an object that is normally at a cooler temperature (T_c) but which is at a higher temperature (T_h) for a fraction f of the time. The result for net lifetime is

(6)

where

There are two situations of interest: objects in cold storage that are occasionally retrieved to room temperature, and collections at room temperature that are occasionally exposed to high temperature. Figure 7 plots the relationship for retrieval from cold storage, and Table 6 provides worked examples.

Figure 7. Effect of Time Out of Cold Storage (Based on Michalski [2000])

There is no advantage to very low temperature cold storage if the object is retrieved frequently to room temperatures. Cold storage does not reverse the decay that progresses during warm periods. Temperatures for cold storage should be designed considering the expected retrieval pattern. Examples in Table 6 can be considered the break-even point, where the retrieval pattern has cut the potential of the cold temperature by one-half. Lower-temperature storage will not significantly improve remaining lifetime, unless retrieval time also diminishes.

Figure 8 plots the loss of lifetime from chemical risk during periods of high temperatures as compared to room temperature (68°F). An annual accumulation of about 35 days at 100°F cuts lifetimes at 68°F in half. Using high temperatures (e.g., 140°F for pest control should not exceed a total of 6 hours per year (or 60 hours each 10 years) to maintain 90% of normal lifetime; this is a reasonable trade-off for reducing the risk of massive insect damage.

A seasonal swing in temperature can allow energy savings, especially in climates with cold winters. Its benefit to collections with low chemical stability is much more important than the smaller risks from annual temperature fluctuation. Because of the exponential dependence of lifetime on temperature, summer adjustments must be balanced by even larger winter adjustments. If one assumes a typical annual schedule (Figure 9) of two winter months at the lowest temperature, two summer months at the highest temperature, and four months of adjustments through the two swing seasons, then the correction between the midpoint temperature and the effective annual temperature for calculating lifetime is given by Figure 10. The correction alters slightly with the midpoint temperature (the solid lines are for near 32°F, and the dashed lines for near 68°F), but for cautious estimates at any temperature, use the solid lines. Table 7 provides several worked scenarios. For example, when the seasonal adjustment reaches ±10°F or more, almost all aging occurs during the summer months, and winter simply becomes a dormant period in comparison.

Figure 8. Reduced Lifetime Caused by Occasional Hot Conditions (Based on Michalski [2000])

Figure 9. Seasonal Patterns Used for Sudden and Gradual Changes

Figure 10. Correction to Temperature Midpoint Caused by Seasonal Adjustment (based on Taylor, forthcoming)

Retrieval from cold storage raises the question of whether to build a transition space, and what procedures to use for acclimatization. Two risks are mitigated by a transition space: condensation during retrieval (the major risk) and direct mechanical effects of the temperature change (usually minor). A transition space adds complexity to cold storage construction and operation. In smaller installations, it also represents a large fraction of “lost” storage. The greatest risk of condensation is during reentry to warm conditions or during failure of the cooling system, so it is essential that objects in cold storage always be inside moisture proof packaging or bags, and that these packages not be opened until the object has reached room temperature. This packaging reduces the need for tight control of relative humidity fluctuations in cold storage, because response times are many days or weeks (see Table 8). When moving from extreme cold storage (–4°F), small amounts of condensation can still form inside packages, such as film cans (Padfield 2002) and larger wrapped plastic objects, causing irreversible blanching of some plastics (Shashoua 2004, 2005, 2008) Despite these side effects, cold storage remains the only option for preserving low-chemical-stability materials (Shashoua 2014). Detailed advice for retrieval of paper, film, and magnetic media is available in the ISO standards listed in Table 4.

Approximate thermal response times of objects are familiar from common experience, and their calculation is described in Chapter 25 of the 2017 ASHRAE Handbook—Fundamentals. This section focuses on the response time that is much less familiar to most engineers, but much more important to preservation science: hygric response times.

Table 8 provides hygric half-times for common objects and enclosed objects in cultural institutions. Where direct measurements are available, these have been cited; other estimates are based on calculations. Lower temperatures greatly increase all these times (Adelstein et al. 1997); conversely, higher temperatures shorten them. Response times at least double for each drop of 18°F: for example, ×2 for cool conditions, ×4 for cold, ×8 or more for freezing storage.

The hygric half time of a plate (a shape that applies to many cultural objects, or that provides an upper bound to cylinders, cubes, spheres, etc) is given by Crank (1979) as

(7)

where

Some estimates in Table 8 are based on Equation (7) and diffusion coefficients found in the literature on polymers. An object with a surrounding moisture barrier (coating or enclosure) can be simplified as a series of two resistances to moisture flow. Figure 11 was generated using this approach, with wood data from Siau (2012) for medium-density wood near 50% rh. When barriers provide useful resistance (upper lines in each plot in Figure 11), the plots have a slope of 1 (linear). When coatings provide negligible resistance (e.g., the boundary layer of air in a calm room), the slope changes to the square law of Equation (7). There is only a slight curve as thickness drops to 0.04 in. across the grain. As Kupczak et al. (2018b) show, the contribution of the boundary layer of air, though measurable, is far from rate determining even for a single sheet of paper. In practical terms, although thinner objects (e.g., a violin, ivory miniature, paper sheet) respond more quickly, they benefit greatly from even simple coatings or enclosures, whereas massive objects do not.

Figure 12 was generated using the same series resistances model, but using leakage equations for an enclosure found in Michalski (1994). The horizontal plateaus are in the region where crack leakage is insignificant, and the wood coating dominates halftime. The slopes are the region where leakage (infiltration) dominates, and the knees indicate which size cracks are worth blocking.

Figure 11. Calculated Humidity Response Times of Wooden Artifacts

Figure 12. Interaction of Air Leakage, Wood Coating, and Textile Buffering on Response of Wooden Chest of Drawers

The humidity half-time of a leaky enclosure with a hygroscopic material (objects or additional buffers) can be expressed in terms of the fraction of the enclosure volume filled with the buffering material (Michalski 1994):

(8)

where

The critical role of leakage N for enclosed objects can be seen in Table 8 for paintings or works on paper in a sealed glass frame (increasingly used by major galleries, especially for loaned paintings.) Depending on tiny differences in crack width, performance can change by orders of magnitude. This is because of the key role of infiltration in determining N, and the fact that infiltration varies with the cube of crack width (laminar flow) (Michalski 1994).

Equation (8) can be reduced to an estimate for materials such as paper, wood, leather, and dense fabrics near room temperature (68°F), where C_ws= 0.00108 lb/ft³. Using a conservative density of ρ ≈ 37 lb/ft³ and a conservative hygric capacity of α ≈ 0.05, then

(9)

A leakage rate of 1 air change per day (acd) is considered a suitable design target for airtight museum display cases (Thickett et al. 2007), so an enclosure half full of wood or paper (V_h/V_e = 1/2) would give a half-time of 600 days. In practical terms, very tight enclosures are rarely very full enclosures (although wrapping large wooden objects in heavy-gage, perfectly sealed polyethylene can achieve this). Half-full enclosures are generally cabinets or crates, and they leak closer to 1 air change per hour (ach), which still provides a 25 day half-time. A tight display case of 1 ach rarely has more than 10% of its volume filled with hygroscopic material, resulting in half-times up to 120 days. In practice, it is difficult but not impossible to design and maintain very low infiltration.

An inevitable concern with enclosures is the humidity fluctuation driven by a thermal fluctuation. This worry first emerged in the 1960s for works of art in shipping crates, but Toishi (1959) showed that if a sealed crate contained hygroscopic material, it would stabilize its relative humidity despite drops from room temperature to freezing and back. Stolow (1966) provided complete data and equations for the counter-intuitive finding that, with natural hygroscopic materials, case humidity even drops slightly when temperature drops (the opposite of empty enclosures) because of the slight downward shift of moisture isotherms at lower temperature. Thomson (1964) showed that the transition point between relative-humidity-controlled enclosures and empty-enclosure behavior for hygroscopic materials such as wood occurred at about 0.6 lb/ft³, and humidity control was fully in place by 6 lb/ft³ (about 2% full by volume). Later authors examined further side effects such as mixed thermal/hygric dimensional response in wood (Richard 2007) and condensation in air pockets during cold storage retrieval (Padfield 2002; Shashoua, 2005, 2008) but the general consensus is that such occasional side effects do not outweigh the benefits (Richard 2007; Shashoua 2014).

From the outdoors, different-sized particles can infiltrate museums. Gaseous outdoor pollutants such as nitrogen dioxide and ozone can also penetrate buildings, including modern HVAC-equipped construction, when gas filtration is not present to remove them. Within buildings, sources of airborne pollutants include institutional activities such as food preparation, service vehicles in the loading dock, and renovation (e.g., preparation of new exhibitions). Construction products such as wood, paints, adhesives, and sealants, especially by-products formed by chemical curing or solvent release, can be important sources of gaseous pollutants. Collections themselves can be sources of pollutants that can affect other objects nearby; examples include archival materials such as cellulose nitrate and acetate films, as well as acidic papers. Collections made of natural organic materials such as leathers, fur, and wood elements can also release harmful volatile compounds. The metabolism of staff and visitors further contributes to airborne pollutants, and introduces coarse particles from skin cell shedding and clothing. It is important to understand that the impact of gaseous pollutants varies according to the sensitivity of each material (i.e., acetic acid corrodes lead but is harmless to silver; silver is very sensitive to hydrogen sulfide, but lead is minimally affected). In other words, the potential damage is very specific to each pollutant/material system and the damage caused to objects by pollutants is usually cumulative, irreversible, and disfiguring. Table 9 presents a list of objects sensitive to various pollutants and pollutant sources.

Dust deposition is a general problem for all collections. More precisely, dust deposition impacts objects’ aesthetic appearance and affects conservation considerations (e.g., cleaning frequency, risk involved during treatments). Coarse dust is relatively easy to remove from robust surfaces such as flat glasses or metals, but difficult to remove from fragile surfaces such as feathers. Particles generated by people are not usually removed by HVAC filters. Fine particles such as black soot pose a particular challenge. This is a typical problem for museums in the vicinity of high-volume diesel vehicle traffic. Special conservation skills are needed to remedy this situation, but there are cases where the soot cannot be removed (e.g., soot entrenched in the cracks of an ivory sculpture, soot in fragile textiles that may be significantly physically damaged by cleaning treatment).

The deterioration process caused by airborne pollutants can be enhanced in the presence of water vapor. This has a relevant impact on objects both in a direct and indirect way. In the presence of high relative humidity, many processes of deterioration accelerate. An example of an indirect effect that is greatly affected by high relative humidity, especially above 75% rh, is the increase in corrosion rates of many metals by pollutants. For example, formaldehyde does not corrode lead at 75% rh, but corrosion can occur at higher humidities (Thickett 1997). Water vapor can directly affect some materials (e.g., cellulose papers, cellulose acetate, nitrate plastic films) by hydrolysis. With just moisture in the cellulose, deterioration is slow, but increases when acids are present. This reaction is called acid catalyzed hydrolysis. Over time, the acids present as by-products of cellulose degradation increase, which further speeds up the reaction (Dupont et al. 2007; Zou et al. 1996). To maximize preservation of objects affected by pollutants, it is usually better to keep relative humidity low, particularly for metal objects. However, the environment must be compatible with the appropriate humidity range established for preservation of organic or composite collections. Oxygen in the indoor environment also may react with objects. Natural rubber is particularly known to degrade by oxidation, and many colorants are vulnerable to fading in the presence of oxygen and light.

Some work has been done to quantify the impact of pollutants on various materials, based on the concept of the lowest observable adverse effect dose (LOAED). This dose is derived using the reciprocity principle: if a critical adverse effect is observed on an object after 1 month at 1000 parts per billion (ppb) of a pollutant, the same damage could occur after 10 months at 100 ppb. When extensive data exist for a pollutant/material system, a no observable adverse effect level (NOAEL) can be determined with some confidence. After studying the effect of acetic acid on (untarnished and pure) lead at different concentrations and relative humidity levels for a year, Tétreault et al. (1998) established a NOAEL for the acetic acid/lead system at 430 μg/m³ or 170 ppb. Extensive sets of LOAED and some NOAEL data have been compiled by Tétreault (2003).

Note that concentrations of gaseous pollutants can be reported in either volumetric units (ppb), which are temperature and pressure dependent, or in gravimetric units (μg/m³), which are temperature and pressure independent. To standardize reporting for volumetric units, the Compressed Air and Gas Institute (CAGI 2012) recommends using standard conditions of 68°F and 14.5 psi. IAQ in Museums and Archives (IAQ 2016) provides an online concentration converter for major pollutants.

In general, there are three scenarios where objects can be at risk in museums:

More information on the issue of pollutants in museums and historical buildings can be found in Anaf et al. (2015), Bellan et al. (2000), Bonacina et al. (2015), Grau-Bové and Strlic (2013), Grzywacz (2006), Hatchfield (2002), Lloyd et al. (2007), Mleczkowska et al. (2016, 2017), Nazaroff et al. (1993), Paterakis (2016), Pretzel (2003), and Tétreault (2003, 2017, 2018).

Silver is very sensitive to reduced sulfide compounds, mainly hydrogen sulfide (H₂S) and, to some extent, carbonyl sulfide (COS). Sulfur sources are many: the outdoors, from people in the room, and from products and collections inside enclosures. It is usually best to keep silver objects in airtight enclosures with no sulfur-emitting products. Consult the combustion section in the safety material sheet (SMS) for specific products to see if they contain sulfur compounds; products that contain sulfur compounds should be avoided. It is also wise to confirm the absence of sulfur compounds in the product by running a spot test, such as the lead acetate test, Oddy test, or equivalent (Robinet and Thickett 2003; Tétreault 2003). The same strategy can be applied for the preservation of copper. The LOAED for H₂S for silver is 0.10 μg/(m³ · yr) (0.071 ppb/yr) and 1.0 μg/(m³ · yr) (0.71 ppb/yr) for copper (Tétreault 2003). Complete dryness will not stop tarnishing, but will minimize it.

The most harmful vapor to lead is acetic acid. Lead is not usually at risk of corrosion in a room but may be in enclosures. Any organic-acid-emitting products or objects should be avoided. Lead may never be safe in the presence of wood, painted wood products, or freshly applied sealants or adhesives. The worst situation would be having lead present in a freshly painted enclosure with paint formed by oxidative polymerization (e.g., oil based paint). Polymerization releases aldehydes, organic acids, and peroxides. Those peroxides can convert aldehydes into organic acids (Raychaudhuri and Brimblecombe 2000; Tétreault 2011). Enclosing lead objects in a display case freshly sealed with acetoxy-cured silicone also puts the lead at high risk of corrosion. A relative humidity kept below 35% prevents corrosion by organic acids above the NOAEL (170 ppb).

Calcareous objects (e.g., limestone, ceramics, shells) can react with organic acid vapors, especially when contaminated by chloride or nitrate salts (Halsberghe et al. 2005) in highly humid environments. No data exist to quantitatively assess these objects’ vulnerability. As a precaution, it is best to minimize the presence of acid-emitting products or objects in the enclosure as well as relative humidity and temperature fluctuations, and if possible, lower the relative humidity to prevent salt dissolution, reaction, and migration.

Some historical glasses degrade slowly in the presence of water vapor, resulting in alkali leaching, which can form crystalline corrosion compounds on the surface or modify the structure of the glass. The presence of formic and acetic acids accelerate the leaching (Robinet 2006). These types of glass should be displayed or stored near 40% rh with very little fluctuation (Koob 2006). See van Giffen et al. (2018) for detailed recommendations on climate control for glass. Enclosures should not contain products that can emit organic acids.

Many colorants (organic pigments and dyes) are known to be sensitive to photooxidation and/or to hydrolysis (Reilly 1998). In addition, some colorants are affected by gaseous pollutants. The most sensitive colorants to nitrogen dioxide, sulfur dioxide, and ozone are curcumin, dragon’s blood, aigani, realgar, iron ink, enju, basic fuschin, Brilliant green, pararosaniline, indigo, madder lake, Persian lake, and saffron (Cass et al. 1989; Whitmore and Cass 1989; Williams et al. 1993). Yellow dyes from photographic prints have been found to be affected by acetic acid (Fenech et al. 2010). Artworks with vulnerable colorants should not be displayed long term without protective enclosures, and photograph prints should not be enclosed with products that may release organic acids.

For many decades, sulfur dioxide was thought to be the most damaging pollutant for paper. As its concentration in the environment decreased over the years, it was found that nitrogen dioxide was the main problem for paper. Fine particles and ozone also affect unprotected paper (Bartl et al. 2015; Gurnagul and Zou 1994). At the room level, displaying art on paper without protection (e.g., glazed framing, display cases) is not recommended. However, formic and acetic acid emitted by various organic materials can affect cellulose, but in the presence of aldehydes, the damage is found to be reduced (Tétreault et al. 2013). As a precaution, however, avoid acid-emitting products.

For paper in books, most damage (yellowing, embrittlement) by outdoor and indoor pollutants tends to remain on the margins of the paper sheets, with very slow diffusion into the book. Many archivists will accept some limited deterioration of the pages’ edges. If stack or single-sheet papers are framed or protected in airtight boxes, gas filtration in archives and libraries may not be required. The cellulose is best preserved against acid-catalyzed hydrolysis by keeping the relative humidity and temperature as low as possible.

Cellulose acetate films degrade by acid-catalyzed hydrolysis, and acetic acid is the by-product released (Reilly 1993). It is best to preserve films from the 1950s and 1960s in cool or cold rooms (see Table 13B). In ambient conditions, degraded films should ideally be stored in special ventilated cabinets to avoid the risk of damage to other collections. Otherwise, consider enclosing the films in airtight enclosures with moisture sorbents to prevent the ingress of high humidity in the storage area (Nishimura 2015).

As with any cellulosic material, cellulose nitrate (CN) films degrade by acid-catalyzed hydrolysis, releasing nitrogen oxides. Old CN films, produced mainly from 1896 to 1952, are unstable and must be kept absolutely below 100°F, above which there is a high risk of self-ignition. CN films should be removed from the collection and properly stored according to NFPA Standard 40, which provides detailed information on the ventilation requirements. However, it is best to preserve these films in cold rooms (see Table 13B).

Other CN objects (such as faux tortoise shell) do not degrade to the same magnitude as films, but to avoid the risk of damage from nitrogen oxide emissions to other collections, CN objects should be stored either in well-ventilated rooms or in special ventilated cabinets (Coughlin and Seeger 2008). A room with a high load of CN items must also comply with local regulations for explosive and combustible substances.

All objects are susceptible to particle deposition, but cleaning of particles is difficult or even impossible for some objects. During handling and cleaning, there is also a risk of physical damage. Example objects include those with powdery pigments or surfaces (e.g., some painted ethnographic objects, butterfly wings); physically fragile objects (e.g., insect collections, filamentous mineral specimens); objects in which fine particles could become lodged in microcracks or interstices (e.g., ivories, painted objects with cracks); and objects with sticky surfaces (e.g., some deteriorated plastics, some polyethylene-glycol-treated wooden waterlogged objects). For these objects, it is best to display and store them in airtight enclosures or in cases with a positive-pressure system. If enclosure is not an option, it is recommended to maintain a minimum distance between visitors and fragile objects: for example, a distance of 5 to 6.5 ft reduces dust deposition by 50 to 75% (Lloyd et al. 2007). This distance prevents deposition of coarse particles on objects, but has limited effect on fine particles because of their longer suspension time.

Introducing outdoor air into interior environments typically increases sensible and latent system loads and serves as the primary source of particulate and gaseous filtration loads. In collections spaces, where the primary goal is maintaining the appropriate interior temperature and relative humidity conditions, outdoor air quantities should be restricted to the minimum necessary for occupancy based on local code. For nonoccupied collections spaces and spaces with periods of zero and nonpeak occupancy (e.g., galleries, reading rooms, workspaces during closed periods), designs should incorporate means of further reducing outdoor air volumes, even to fully closed, based on actual need. CO₂ sensors and modulating dampers can help automate this process, and may allow for flexibility with certain code requirements. Particular consideration of outdoor air requirements should be given to spaces housing materials that may emit hazardous substances (e.g., radon) or require specific outdoor air volumes because of fire code.

Economizer controls, typically intended for energy-savings/free-cooling of interior environments, should generally be avoided in cultural heritage applications. Dry-bulb temperature and enthalpy controls may allow inappropriate levels of moisture (either too wet or too dry) into the airstream, increasing latent loads for dehumidification and humidification compared to the return airstream. Although dry-bulb temperature and dew-point controls can be programmed to allowable conditions, they generally offer reduced energy benefit compared to other energy-reduction strategies because, generally, most outdoor environments align with both temperature and dew-point requirements only for short periods. Economizers also increase risks to interior environment maintenance: control failure or mechanical failure of outdoor air dampers on ductwork sized to allow for 100% of the system volume can quickly create significant environmental issues. Air-side economizers should not be used unless (1) bin analysis or other study shows outdoor air moisture content to be favorable for an economical number of hours, and (2) favorable outdoor air can be reliably selected by the control system by combined dry-bulb temperature and dew point comparisons.

Positive air pressurization has been frequently used in cultural heritage facilities to minimize incursion of external loads into controlled collections environments. However, this practice often increases energy consumption, and in some cases increases the risk to the building envelope, particularly in historic structures. With improved envelope design and appropriate zoning and adjacency design, positive air pressurization is no longer an absolute requirement in cultural heritage settings. Neutral pressurization is typically an appropriate goal; avoid negative air pressurization. In multiple-story buildings, stack effect (discussed later) creates unavoidable pressurization in upper floors; airflow design should not exacerbate this problem. Positive pressurization is typically created through a combination of outdoor air and duct design, with supply air ducts sized for greater volumes than the return air. Designs for pressurization in cultural heritage facilities should allow for equal volumes of supply and return air to the downstream zone to facilitate recirculation modes (no outdoor air) without pressurization consequences because of duct sizing. Modulating dampers on the outdoor and return airstreams as well as using adjustable return and supply air grilles can allow for balancing adjustments.

In some circumstances, natural ventilation may be necessary for interior moisture control and/or inhibition of mold growth. Historic structures with limited mechanical intervention may benefit from controlled natural ventilation on a scheduled basis (e.g., diurnal or seasonal operation) or may require either mechanized or passive ventilation for emergency situations or disaster recovery. The goal of ventilation is one of the following:

These operations may be enabled by a high-limit relative humidity sensor in the space, time scheduled, or (for disaster recovery or power outages) manually activated.

Air change rates in cultural heritage institutions are not constant values, and should vary based on zone usage and occupancy, with other specific factors (e.g., events spaces, fabrication or paint shops, conservation labs, off-gassing collections materials) accounted for as necessary. The operational goal after optimization is to run the system with the minimum air volume/change rates necessary to maintain the desired environmental conditions while providing for occupancy and protecting against microenvironments. Proper envelope, airflow design, and duct layout should minimize potential microenvironments; environmental data logging in conjunction with control sensors can alert staff to potential issues. Initial design may use air change rates recommended for particular zone types (office, laboratory, classrooms, etc.) but should include variable-frequency drives (VFDs) or variable speed drives (VSDs) that can control air volume/change rates based on occupancy patterns, established needs, and other factors. Collections storage zones generally require lower air change rates unless extenuating factors (e.g., off-gassing materials, issues with microenvironments) dictate otherwise.

Stack effect can have significant implications for control in cultural heritage settings, particularly in multistory structures and high-ceilinged spaces (ranging from modern high-bay storage environments to historic structures that may incorporate historic frescoes and murals). Differences in temperature (and, to a lesser extent, moisture) between interior and exterior environments can result in density gradients that induce air movement and exchange, drawing unconditioned outdoor air into the building and often causing issues with airflow, microenvironments, and overall system operation. Stack effect may reverse depending on the exterior and interior conditions: when cooling indoors, upper areas may be negatively pressurized relative to outdoors, drawing warm air into the structure, while lower parts of the building may be positively pressurized. The reverse is true when heating indoors. This is particularly problematic in structures with limited envelope integrity and limited or poor zone design. Where this effect is noted, if envelope improvements are not an option, pay particular attention to airflow design and balancing to combat preservation risks (typically from high-temperature and high-relative-humidity microenvironments).

Interior thermal stratification can occur even in buildings with excellent envelope integrity. It is caused by the displacement of warm air by more dense cool air. This can occur independently of interior/exterior pressure differentials and, like stack effect, can create issues in preservation environments because of microenvironments and poor environmental controls throughout multistory or high-ceilinged spaces. Common problems include high temperatures (which can increase rates of chemical decay) near ceilings and on upper levels, and issues with high relative humidity and potential mold risk because of cooler temperatures near floor level, especially in areas with poor air circulation. Proper zoning, airflow design, duct layout, and balancing can reduce stratification. For storage and cool environments, overhead diffusers and floor-level returns generally are preferable; occupied spaces (offices, galleries, etc.) may use floor-level or overhead diffusers. Ceiling or circulation fans may be used as low-impact solutions for improved air mixing and reduced stratification.

This specialized approach has limited application and must include safety controls, but is sometimes the only option that can handle envelope limitations in cold climates. In this approach, the heating system is controlled by a humidistat rather than a thermostat (LaFontaine and Michalski 1984); cold, damp air is heated until the relative humidity drops to a predetermined safe range, typically below mold germination conditions. Where interior temperatures drop consistently below 50°F, it solves the problem of humidity in a building that does not have an adequate envelope. Humidity-controlled heating does not provide human comfort in winter, but many small museums, historic buildings, and reserve collection buildings may be largely unoccupied during this period. A high-limit thermostat is necessary to stop overheating during warm weather, and a low-limit thermostat may be used if water pipe freezing is a concern. This approach has been used in Canada (LaFontaine 1982; Marcon 1987), the United States (Conrad 1994; Kerschner 1992, 2006), and in many historic buildings in Britain. Maekawa and Toledo (2001) successfully applied humidistatic control in hot, humid climates to minimize mold growth.

Some cautions apply. Foundations in a previously heated building may heave if the ground is waterlogged before freezing. Improving drainage, insulating the ground near the footings, and heating the basement reduce this risk. Problems have occurred in buildings with dense object storage and a very low infiltration rate, such as a specially sealed storage space (Padfield and Jensen 1996); a very slow supply of dehumidified air to the space can be helpful.

This approach is cost effective in seasonal museums (especially for low-mass wood-frame buildings) in colder climates such as the northern United States and Canada, and in maritime regions. Application in hot and humid environments should be judiciously considered, typically where mechanical dehumidification is impractical. Humidistatically controlled heating may be applied where the imminent risk of mold growth outweighs other degradation risks, and should be balanced with the increased risk of chemical decay because of elevated temperatures. In many circumstances, improved air circulation or natural ventilation for air circulation may be a preferred first step for mold avoidance.

Note: humidistatically controlled heating may be used in place of stand-alone dehumidifiers. As described in the following section on Dehumidification, stand-alone dehumidifiers and air conditioners pose a particular threat to cultural heritage collections because of the inherent risk of flooding and electrical fire in the local collections zone; their use should be judicious, and only when the building/space is occupied.

Control of mechanical operations in hot and humid environments (whether constant or seasonal) depends largely on the level of mechanical intervention selected. From a control perspective, moisture management (both relative humidity and moisture content/humidity ratio) is the critical process, and may be achieved through various mechanical means, including dehumidification, cooling with secondary dehumidification (as in typical direct expansion/refrigerant-based window, residential, and package air conditioners), ventilation, and, less commonly, humidistatically controlled heating. In most applications, the primary preservation goal is to restrict mold growth and other biological risks, with mechanical damage (particularly in seasonally humid/dry climates) and chemical decay typically secondary concerns.

Envelope capability heavily influences both control design and equipment selection. Where the structure has a modern, purpose-built envelope that can limit sensible and latent loads, temperature control and the limitation of chemical decay may be the first design priority. For most historic, renovated, and/or repurposed structures, control design for collections zones should primarily be based on space relative humidity. Designs should consider a specific ventilation control (whether integrated into the primary system, or as a separate system) that can also be manually enabled in the event of limited power availability or long-term power outages, where generator capacity may only be capable of providing circulation without temperature or humidity control. In mechanical designs where redundancy or back-up power may not be available, control and system designers should also consider advocating for passive strategies, including single-side, cross, or stack ventilation, as a way to provide airflow and limit mold growth during equipment failures or power outages.

Additional information on environmental management in hot and humid climates is presented by Harriman (2009) and Maekawa et al. (2015).

Selection. Selection of temperature and relative humidity sensors (thermostats, humidistats) should consider accuracy, initial calibration, and response time. Sensors for cultural heritage applications generally trend toward the more accurate, with ranges of ±0.36°F and ±2% rh or better. As a guideline, response time should be within 1 min for temperature, and less than 2 min for relative humidity (BSI 2010). Reducing project costs by reducing relative humidity sensor reliability and accuracy is a false economy that will compromise operational accuracy and place the collection at risk. Dew-point sensors, though initially expensive, are much less likely to suffer from calibration drift.

Calibration. Design and initial commissioning typically assume newly purchased hardware and represent the best-case operational accuracy for the system. Institutional staff should be provided with guidance for eventual replacement or recalibration to provide continued accuracy in operation. Note that field recalibration rarely achieves the same level of accuracy as factory recalibration (if available) or new equipment.

Location. Typical sensor placement is based on control points, with occasional reference data (e.g., discharge air conditions, space or return air conditions, outdoor air conditions). Generally, primary control sensors should be located in the collections spaces; each collections space should have at least one sensor associated with the control system. As environmental optimization in cultural heritage has increased in popularity, institutional staff may desire greater transparency in operation, with data available from multiple points to assess the performance of individual system components. Sensor locations may now consist of an expanded list of reference points, beyond typical control points, and may include

Designers should work with institutional staff to understand future informational needs and determine appropriate sensor locations.

Stand-Alone Data Loggers. Many cultural heritage institutions use stand-alone digital data loggers to monitor a collection’s environment for preservation purposes. These devices have the advantage of flexible deployment throughout a mechanical zone (close to the collection, or in areas of suspected microenvironments) compared to hard-wired control sensors, which, because of access, location, calibration, etc., may not always provide an accurate representation of what the collection experiences. Stand-alone data loggers are often equally as accurate as BAS/BMS sensors and may have the advantage of more regular calibration. Use of data loggers and data comparison should be discussed during predesign; ideally, data from both systems (data loggers and BAS/BMS) should be used to assess environmental performance and to identify potential issues with either set of equipment. Some institutions may still use hygrothermographs to record environmental data; however, without continuous maintenance and frequent calibration by trained personnel, these units are prone to large measurement errors and are generally no longer recommended for use in the cultural heritage field.

CO₂ Sensors. Consider using CO₂ sensors as a control mechanism in settings where there are opportunities to control outdoor air intake beyond a set volume. In storage environments and collections spaces that are only lightly occupied or that are occupied on a fixed schedule, there may be significant opportunities to reduce heat, moisture, and filtration loads by minimizing outdoor air quantities when fresh air requirements are flexible.

Now common for both fans and pumps, VFDs or VSDs should be included on most systems larger than residential/light commercial equipment; small drives are regularly found on rooftop and other package systems. As control equipment, their uses vary; drives may respond to differential pressure or to downstream damper control in a variable-air-volume system or, more critical in collections-centric zones, can be used for optimal part-load fan operation or soft stops/starts as part of programmed shutdowns.

During predesign, projects for existing buildings should consider an energy audit based, at minimum, on existing systems and their typical operation. Predesign teams may also consider comprehensive energy audits for the entire building, to identify the influence of additional energy factors (e.g., lighting) not addressed in a mechanical study. Data from the audit should be analyzed for evidence of excessive operation, with these findings informing future system design. After construction, institutions may perform periodic energy audits to inform continuous commissioning or optimization processes.

Applied to mechanical systems, LCCA is the assessment of the whole cost of the potential installed system over its estimated lifespan. This exercise, which can be applied on a whole-building scale, greatly assists in selecting the most appropriate design solution when several options appear viable. Typically, LCCA for systems should include estimates of

LCCA studies should carefully identify the variables being compared, especially regarding differences in systems versus differences in environmental conditions. For example, comparative analysis of potential system designs that achieve the same environmental conditions (e.g., water-based subcool/reheat versus direct expansion cooling and electric reheat) is different from comparing LCCA for a water-based subcool/reheat system designed for either 68°F/50% rh or a 65°F/40% rh.

Energy efficiency may be achieved either through direct equipment efficiency (e.g., more efficient coils or compressors) or through operational efficiencies, such as airflow control, outdoor air control, and strategies for nonpeak operation. In general, system designs should include the most efficient equipment selections possible given budget and institutional infrastructure. Potential operational efficiencies, which are highly dependent on performance and use of the final populated space, should be accounted for through flexible operational design (e.g., VFDs, modulating dampers and valves, programmable thermostats in smaller applications) and tested during continuous commissioning or optimization work before adoption.

Lighting and daylighting, as they pertain to system design and operation, have several components to consider. Light-emitting diodes (LEDs) are increasingly common choices for collections environments in both new construction and renovation, and unlike other lights (fluorescent, incandescent, halogen) have comparatively little impact on overall system operation and design. Although LED fixtures do produce heat, the quantities are considerably smaller. Engineers should consult with lighting designers regarding lighting choices, their potential heat output, and what load must be accounted for in design calculations. Renovations of systems serving existing environments using a variety of heat sources should consider a lighting audit to determine existing thermal loads from space or exhibit lighting.

Daylighting is often proposed, especially for noncollections areas. Generally, daylighting should be avoided for all storage environments and avoided, or severely limited, for most exhibition spaces. Preservation assessment for possible light impact on collections should be conducted in conjunction with any lighting design. Ayres et al. (1990) noted that daylighting is always a net energy penalty. If used, the daylighting aperture should be minimized, and avoided as much as possible in and over collection areas. For lower risk of leaks and better-managed lighting, clerestories are preferred over skylights. In applications where daylighting is unavoidable (e.g., historic structures), light-reducing and blackout shades and UV filters can reduce exposure of collections to both visible light and ultraviolet wavelengths, as well as reduce potential heat gain.

As detailed in Chapter 6 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment, hybrid HVAC systems use multiple means of heat transfer to control a specific environment. Most commonly, this involves collocating radiant and forced-convection systems. Most of the sensible loads are assigned to radiant panel systems, whereas latent loads and the remaining sensible loads are assigned to forced-convection systems. Decoupling the HVAC functions primarily into sensible and latent heat transfer components allows the designer to select better function-oriented HVAC components and ensure higher accuracy and precision in control, potentially at lower energy expenditure. Application has been limited in collections environments; where the approach is used, in-floor radiant may be the most appropriate application because of the potential for microclimates against radiant wall panels in exhibition and storage environments. One key detractor for application in cultural heritage is the possible higher cost of installation and maintenance of dual systems.

Historic structures, house museums, and recent buildings can present a similar scenario from a legacy perspective when institutions seek to combine modern forced-air cooling and dehumidification with original hydronic systems using radiators or a perimeter heat loop for either a portion or the whole of the sensible heating load. The heat transfer mode remains the same (both systems operate by convection), but the operational challenge is similar to that with other hybrid heating system designs: management and control of a hydronic system in conjunction with forced-air control. Balance and efficient operation can be excessively difficult to achieve; new construction, with opportunities for insulated glazing and perimeter forced air zones, can generally avoid using dual hydronic and forced-air systems.

Dual-fuel and multiple-energy-source systems are typically seen in residential or light-commercial applications, often as a package unit that incorporates both fuel and electric heat depending on demand. Application of these units in cultural heritage is relatively limited, but as new technologies develop, other multiple-energy-source systems have begun to emerge, particularly for smaller applications and heating/dehumidification side. Water heating may use recovered waste heat from various applications or use solar thermal collectors, with a traditional boiler for back-up; the hot water can then be used for heating, reheat for dehumidification, or regeneration for a desiccant system. Variants on hybrid designs, especially with renewable energy sources, may be particularly applicable in design situations where regular utilities are limited (e.g., well-water or tank fuel-storage systems) or the cost of certain forms of energy is prohibitively high.

As discussed previously, maintenance, accessibility, and ease of operation and repair are critical components of any proposed mechanical design. System design should enable both preventive and reactive (i.e., breakdown) maintenance by either existing institutional staff or an identified contractor in the local area. Specialized equipment with no local maintenance or repair support should be carefully considered for potential benefit versus risk because of lack of maintenance, and alternative options should be explored. For large projects in larger cities, code-required staffing for the plant should be considered; sometimes, smaller reciprocating chillers can be used at night to preclude the licensed engineer needed to operate larger chillers. Specific aspects to consider as part of design review may include

Effects of maintenance activities on the collection must always be considered. For example, testing or accidental activation of a smoke removal system can radically change the collection environment. Tools and ladders in gallery and storage areas are a threat to the collection, and special precautions must be taken. Contaminated air conveyance components (e.g., mold growth and other build-up) can contribute to pollution levels, lead to premature component failure, and affect heat transfer efficiency, resulting in higher utility costs. Regular inspection and cleaning are an important part of preventive maintenance.

As discussed in the section on Zoning (under Controls Design), three general environmental zones must be considered not only for control but also for system layout and design: noncollections spaces, occupied collections spaces, and unoccupied collection storage environments. Efficient operation is typically best achieved when each of these zones is served by mechanical systems dedicated to that particular environment; this approach avoids unnecessary energy usage (e.g., significant dehumidification for a noncollections zone) and improves the likelihood of maintaining environmental control.

The design engineer should be involved from the start of project planning to ensure that space layout does not present unnecessary problems. For both functional organization and design efficiency, spaces in similar environmental zones should be kept physically contiguous, with the occupied collections zone and the unoccupied storage environments adjacent to one another. This allows not only for efficient ducting, but also for efficient interior wall (thermal and vapor barrier) construction. Appropriate zoning should minimize the necessity for downstream subzone equipment (VAV/reheat, subzone humidification) in most occupied and unoccupied collections zones.

Activities that pose a potential threat to collections environments (food areas, loading docks, fabrication and rough shops, maintenance and housekeeping, etc.) should be kept away from collections zones, with exhaust from these areas located away and downwind from any fresh air intakes for collections systems and the rest of the building. One key exception to this practice may be relative-humidity-controlled crate storage, which may share a zone with occupied collections areas. Loading docks should ideally be positively pressurized, and shared walls with collections zones should be carefully designed to mitigate any potential energy transfer, whether from a load perspective or to reduce risk of condensation or other issues. Where possible, a separate loading dock may be specified for cleaner transfer of collections materials and crates, with direct access to crate storage and collections intake work areas.

System designs that call for interior moisture control, whether humidification or dehumidification, must carefully consider the likely performance of the existing or designed envelope for both thermal insulation and vapor transfer. These issues are commonly discussed for exterior envelopes, but the same issues can exist with excessive thermal or vapor differentials across interior walls.

Thermal insulation/barriers should be adequate to moderate heat transfer through the structure, with particular attention paid to thermal bridges (e.g., wall studs, floor/wall junctions, roof/wall junctions) that may be sources of heat gain/loss or potential condensation points.

Vapor incursion or loss occurs by two primary paths: air leakage/gaps and diffusion. Air leakage or gaps speed heat transfer, but may be defeated for thermal loads with positive pressurization. Vapor carried by air cannot be fully mitigated pressurization: it slows the process by slowing air transfer, but vapor flow and equilibration between interior and exterior spaces continue independently based on vapor, rather than air, pressure. The differential between interior and exterior vapor pressures also drives diffusion (movement of water vapor through permeable materials) from areas of higher vapor pressure to lower. Diffusion manifests differently depending on exterior climates and interior environments. When exterior moisture conditions are higher than interior conditions, whether typically or seasonally, vapor may move inward, especially during dehumidification, adding to system load and causing efflorescence or other damage on interior surfaces. During humidification, vapor is typically forced outward through the envelope, and may limit the possible control of the interior environment and damage the structure’s exterior. Diffusion in either direction, combined with interior wall temperatures, can cause significant structural damage ranging from mold growth in warm environments to structural failure in situations of repetitive freeze/thaw cycles.

Windows, doors, and skylights can pose risks as potential condensation points, as well as points of heat transfer. Repetitive condensation on doors and windows places that element at risk, and the potential transfer of that moisture to the wall or floor can lead to issues ranging from mold and wood rot to cracking and masonry damage because of freeze/thaw.

Older structures, especially those with historic significance, can be particularly problematic, especially in projects that propose indoor moisture control. Where envelope performance is in question, the first step is to reassess environmental goals: considering either higher or lower relative humidity ranges to minimize differences in vapor pressure or, in cold climates, lowering temperature to raise relative humidity rather than using mechanical humidification. If specific conditions are necessary for long-term preservation, alternatives for containment of preservation environments should be considered, such as

Designs should include some way to manage the interior environment during periods of interrupted operation, such as maintenance, equipment failure, power outages, and disasters.

Redundancy and back-up equipment may be applicable in the largest of cultural heritage applications, but capital cost, load cycling for long-term operability, and limited availability of physical space are all factors against redundancy as a typical practice. Rather, discussion commonly turns to reliability and resiliency as risk mitigation strategies for interrupted operation. Better understanding of collection equilibration times to environmental changes and improved building and envelope construction often mean that collections, particularly storage areas, may be able to hold appropriate environmental conditions for longer periods than previously understood. The design team should discuss scenarios, with designs reflecting potential response to events to minimize downtime. Spare equipment may be kept on site to allow for timely repairs, modulating outdoor air dampers may be set to fail closed to minimize infiltration, and in certain environments architectural and systems allowances may be made to provide natural ventilation in long-term disaster or recovery scenarios.

Stand-alone generators may be part of this strategy, but their design and siting should be reviewed carefully for load capacity, accessibility, weight restrictions (for rooftop applications), fuel availability, and locations of fuel storage tanks, which pose their own significant disaster risk.

Accounting for design capacities for both sensible and latent energy is critical in cultural heritage applications and collections environments. Though common when specifying some systems types such as four-pipe subcool/reheat and desiccant designs, adequate latent capacity is often disregarded with two-pipe and standard package unit designs, resulting in environments that may be able to maintain temperature control, but commonly see relative humidity conditions rise beyond safe limits as exterior temperatures climb and cooling capacity is dedicated to the sensible load.

Certain specific load characteristics of cultural heritage buildings should be considered in system design. Occupancy patterns can vary widely among zones. Noncollections zones and occupied collections zones tend to have high occupancy only at certain times, which may include events outside of normal operating hours. Some gallery occupancies are as high as 10 ft² per person, but unoccupied storage zones may have 1000 ft² per person or less. Systems should be designed to handle maximum loads as well as the more common part loads. Many engineers design to 20 ft² per person because part of the room is never occupied. In zones where spaces are extensively used for receptions, openings, and other high-traffic activities, even higher density assumptions may be justified. Continual and close dialogue between the designer and the institution is critical. Where possible, especially in renovation projects, designers should request actual gate counts and event attendance to solidify load estimates.

Lighting loads vary widely by space and by time of day. For some applications, the most common driver of cooling load in a museum is display lighting, particularly where incandescent, halogen, fluorescent, or metal-halide bulbs are still used. The engineer should ensure that estimated lighting loads are realistic. Lighting typically varies from 2 to 8 W/ft² for display areas; figures as high as 15 W/ft² are sometimes requested by lighting designers, but are rarely needed. With growing awareness of damage caused by light, display areas for light-sensitive objects should have low illumination levels and associated low lighting power densities. The proliferation of LEDs has minimized many of the heat load concerns.

Designers should be actively involved in discussions about collections storage designs, which can have significant impact on space airflow and the potential for microclimates. As a general rule, storage solutions fall into four categories for their interaction with the overall room environment:

In all applications, designers must advocate for sufficient spacing for airflow between storage furniture and exterior walls. Poor airflow can lead to high temperature or high relative humidity microenvironments, particularly near exterior walls with inadequate thermal resistance.

As discussed in the section on Response Times of Artifacts, the hygric response time of an enclosure containing hygroscopic materials will almost always exceed 24 h and can reach many months. This is analogous to the thermal mass (flywheel) effect of a building, but with hygric response times typically an order of magnitude longer. Design of sustainable and reliable building humidity control should take advantage of the hygric flywheels filling the building, not fight them. Feeding HVAC supply air into cases, though often proposed, usually results in very erratic conditions inside the cases since the moisture and thermal loads of the cases are a tiny fraction of the room loads. Barrette (1985) describes the failure of a system using AHUs attached to very large display cases at the Metropolitan Museum, and their (successful) decision to adopt “passive” relative humidity control instead.

When the annual average relative humidity of the space is suitable for the object or collection, a completely passive approach is best. Many objects can adapt to a stable relative humidity somewhere between 35 and 60%, and the annual average relative humidity inside many buildings (with HVAC) in many climates is also in this range. The role of the enclosure is to smooth out relative humidity fluctuations. If the only fluctuations are hourly or daily, and the enclosures are more than half full of hygroscopic materials, then ordinary enclosures (e.g., metal storage cabinet full of paper files, photographic materials in impermeable packaging) can perform well enough. If the fluctuations are seasonal, or the enclosure has only a small fraction of its volume filled, then achieving an adequate response time requires special attention to seal details and probably additional humidity buffers (Tétreault and Bégin 2018). The key to achieving long response times is reducing enclosure leakage (Michalski 1994; Thickett et al. 2007). Measuring and specifying acceptable case leakage is well established in museums (Thickett et al. 2007).

When the annual average relative humidity is not acceptable for the object or collection, routine intervention becomes necessary. The most reliable intervention is using a removable humidity buffer, typically silica gel. When the enclosure drifts to an unacceptable relative humidity, the (reusable) buffer is removed and replaced with one that has been equilibrated (reconditioned) to the desired relative humidity. This technique has been used to maintain middle, very low, or very specific relative humidity conditions in display cases (Thomson 1986), shipping crates (Richard et al. 1991), boxes with mineral specimens (Waller 1992), and film packages in cold storage (McCormick-Goodhart 2003). This method provides resilience: if the reconditioning process is abandoned (not unusual), all high risks from incorrect relative humidity are still mitigated as long as the annual average does not exceed 65% rh. Tétreault and Begin (2018) provide a recent manual on the application of silica gel buffers for museums.

Active mechanical control systems for relative humidity alone have been made for museum enclosures, especially display cases. These can be very small commercial units hidden in a single case, or designs for a larger package that feeds many cases via small-diameter tubes (Shiner 2007). Often these are used as short-term solutions to a loan requirement. For permanent collections, experience shows that the key factor for long-term success is that the units have a known parts and maintenance provider, and that they are adopted and maintained by facilities staff, not by the conservation or exhibits staff alone.

Cold or frozen storage vaults extend the life of materials particularly sensitive to chemical deterioration (e.g., cellulose acetate, color photographic materials), or those that require cold/frozen conditions for their own stability as well as the potential threat they pose to other collections materials (e.g., cellulose nitrate). Design environmental conditions (see Table 13B) for cold/frozen environments typically require dehumidification beyond what can be achieved by chilled-water or glycol-mix systems. Direct expansion (DX) systems, though common as the primary equipment for cold/frozen environments in other industries, are typically only used for sensible cooling in these environments for cultural heritage. Dehumidification is handled as a separate process, typically via desiccant equipment, which achieves the preservation requirement for cold temperatures at controlled relative humidity conditions.

Design and construction of frozen vaults, in particular, should be carefully considered from a space need perspective: many cultural heritage institutions have only limited quantities of media that require frozen storage. For smaller collections, storage needs may be met with stand-alone solutions such as frost-free freezers. Depending on intended usage patterns, frozen storage may not be the best solution (see Table 6).

Cellulose nitrate is classified as a hazardous material; beyond preservation requirements, consult NFPA Standard 40-2019 and local fire codes for safe storage and disposal requirements. Smaller quantities can potentially be safely managed with approved cabinetry; vaults that hold larger quantities have specific construction requirements, including fire-rated walls and explosion venting.

Antechambers may be necessary in certain applications, especially where vault temperatures are colder than the typical ambient dew point (whether seasonal or year-round). The antechamber microenvironment should allow the object to fully equilibrate to a temperature higher than the final environment’s dew point, with relative humidity in the safe range for the media. Some collection packaging practices may eliminate the necessity of the antechamber; application varies by institution and should be discussed during predesign.

Detailed discussion of HVAC design for laboratories can be found in Chapter 17. Further details on containment, collection, and removal of airborne contaminants such as particulates, vapors, and hazardous gases can be found in Chapter 33.

Air volume design in most cultural heritage applications should account for varying needs; individual functional zones may have volume requirements that change based on occupancy, space usage, loads, and other factors. In all cases, the design goal should be to provide the appropriate volume required to meet the need, without expending energy on work applied to volumes that are unnecessary. Collections zones should have supply and return air ductwork designed to equal volumes to allow for recirculation without risk of negative pressurization during low-occupancy, minimal outdoor air operation. Diffusers with adjustable outlets should be selected to allow appropriate zone balancing. The specific application of constant-volume and variable-air-volume designs is discussed in the section on System Types.

Figure 16. Basic Components of HVAC System for Museums, Galleries, Archives, and Libraries

Every forced-air convection system, ranging from air-handling units to computer room air-conditioning (CRAC) systems and fan-coil units (FCUs), has at least a supply fan; depending on total air volumes, physical extent of the zone, outdoor air and exhaust design, etc., return fans may also be required. For institutions with the appropriate infrastructure, fan-wall/fan-array designs have distinct advantages over single-fan designs, including ease of motor maintenance/replacement and improved redundancy/resiliency. In either single-fan or fan-wall solutions, each motor should typically be equipped with a VFD; for systems with both supply and return fans, designers should ensure that VFDs are installed on both motors.

Fans may also be required for processes separate from the primary mechanical design. Ceiling and other circulation fans can be very effective for mitigating stratification and microenvironments, and some designs, especially in historic structures, should include fans for natural ventilation or other purposes.

Equipment for sensible heating in cultural heritage applications varies widely; direct-fire solutions are occasionally applied to storage warehouses and as components in desiccant regeneration, and indirect-fire systems, ranging from residential-type systems used in many historic structures to roof-top package systems, are common in smaller applications. Heat pumps may appear in smaller structures in hot environments, and electric heat can be applied in various settings, as the primary heat source or as a downstream component. Coil-based designs are most common in larger cultural heritage applications, and may be served with hot water or steam from boilers or a central plant. Heat generation systems can vary, including steam and oil with converter, modular boilers, and scotch marine boilers. Downstream reheat coils (common in VAV systems, but may exist without) most commonly use hot water or electricity.

Locations of heat application can vary; preheat may be required to protect downstream coils from freezing. For cultural heritage institutions, heat should typically be applied downstream of a cooling coil to allow for dehumidification. Reheat coils as part of downstream VAV configurations are typically best used for increased sensible temperature control. If used as the sole source of reheat, the capacity of downstream in-duct coils should be carefully considered and sized appropriately.

The most common equipment for sensible cooling is coil based, including DX (refrigerant) systems, chilled water, and glycol. Evaporative cooling, common in arid regions, can be difficult to apply in cultural heritage settings because of reduced sensible temperature and moisture control. Designs using evaporative systems should carefully consider performance during any wet/rainy seasons; alternative design or additional equipment may be required for dehumidification. Ice storage systems with glycol may be an option in larger applications with appropriate infrastructure.

Cooling coils are commonly used for both sensible cooling and dehumidification in a subcool/reheat configuration. Location of cooling should typically allow for both sensible temperature control and dehumidification, with the cooling coil located upstream of the heating coil.

Screw compressors are recommended to generate chilled water at 36°F for use in chilled-water coils, which generally have copper fins and tubes.

Humidification should be provided by clean steam or deionized water introduced in the air system. Evaluate the moisture source for risks of pollutants; building/plant steam should typically be avoided as a source for humidification, unless used to generate clean steam via a steam-to-steam heat exchanger. Often, steam used in closed loops to heat is treated with compounds (especially amines) that can pose a risk to the collection (Volent and Baer 1985). Systems should be selected and designed to prevent standing pools of water, and should follow good humidification design (see Chapters 1 and 22 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment).

Humidification equipment varies based on the application. In all applications, deionized or reverse osmosis water (based on institutional needs and availability) are preferred, whether for isothermal or adiabatic systems. Isothermal is generally more common, especially in year-round or seasonally cold environments, and steam is commonly generated via steam-to-steam heat exchangers or electronic steam humidifiers. Adiabatic systems (i.e., evaporative pan humidifiers, spray-coil wetted-element systems, pressurized-water atomizers, ultrasonic humidification) are more common in warmer climates. All materials in humidification equipment should be selected to minimize mold growth and degradation of system components.

Humidification should typically be located downstream of cooling and heating coils, and preferably in the primary air-handling unit. Mechanical design for cultural heritage is often more concerned with controlling humidity than temperature. The averaging effect of a common mixed-return air and common humidifier on a central system is preferred, but downstream zone humidifiers may be necessary to boost relative humidity because of specific environmental requirements or loads. If local low humidity conditions exist, try to identify and correct the cause of the condition before applying further mechanical solutions. Duct installations, whether on a common supply or in downstream subzone applications, should follow design requirements for absorption and be provided with drainage from the ductwork. Downstream humidifiers should not be located above collections areas; if it cannot be avoided, the ductwork approaching and downstream of the humidifier should be fitted with an additional catch pan with a drain and water alarm beneath the duct to provide protection for the collection. Widely different conditions in zones using the same air handler can be difficult to maintain and inefficient. If possible, different zone conditions should have the same absolute moisture content, using zone reheat to modify space humidity for different relative humidity requirements.

Designers should review maintenance requirements for humidifiers with institutional staff or contractors; electronic steam humidifiers and systems using building water can require particularly intensive maintenance to maintain design conditions and remain operable. Humidification in structures with limited envelope capacity should be carefully considered, and typically avoided.

Dehumidification is the single most critical mechanical process for many cultural heritage institutions, and should be a central focus of most system designs. The required environmental conditions often determine the means of dehumidification. Sebor (1995) suggests the following typical approaches to more aggressive dehumidification:

Subcool/reheat designs, typically using cooling and heating coils, can generally achieve dew-point conditions as low as 36 to 38°F, depending on the chilled water/glycol temperature. Lack of dehumidification capacity may originate with design assumptions or equipment selection, or may be caused by issues including compromises in the cooling medium temperature, inadequate reheat, dirty or blocked coils, and poor flow or valve control.

Desiccant dehumidification, most commonly used for low dew points necessary in cold/frozen environments, is now being applied for dew-point control in some applications at 40°F and higher. Desiccant technology (typically a silica-gel rotary wheel design) can be installed as part of the primary system or as a separate component that feeds into the main air handler. For zones with minimal latent loads, or in designs using a high volume/percentage of outdoor air, designers may choose to focus on dehumidification of the outdoor airstream, rather than on the mixed air. For smaller applications and systems, package desiccant systems may be put in line with outdoor or return air. Dehumidification systems should be additions to a typical cooling system; they cannot maintain comfort conditions by themselves. For collections requiring cool, dry conditions, a desiccant system may be required. Chapter 24 of the 2020 ASHRAE Handbook—HVAC Systems and Equipment has further information on desiccants.

Note: stand-alone dehumidifiers, common in residential and smaller applications, are not recommended for use in cultural heritage settings except in emergency scenarios. Risks to collections (e.g., electrical fires, flooding from basin overflow) generally outweigh benefits. If stand-alone dehumidifiers are used in collections environments, it should only be during occupied hours, with units emptied and unplugged before staff leave.

Outdoor air should typically be limited to the minimum necessary in cultural heritage applications. Intakes should have modulating dampers balanced with return air dampers on the system. Siting of the physical intake is critical; potential locations should be evaluated for various factors, including pollutants (e.g., vehicles, equipment, kitchen exhaust), prevailing wind directions, and sources of moisture/particulates (e.g., sprinklers, landscaping). In most applications, outdoor air intakes should fail closed on a shutdown or power outage, with the ability to manually adjust the damper if natural ventilation is required.

Preheat, precooling, or dehumidification may be best applied to the incoming outdoor airstream, increasing operational efficiency.

Ductwork selection varies by application; sheet metal, flexible fiberglass (flex duct), cloth, and others can be appropriate duct choices. Fiberboard is typically not recommended for cultural heritage institutions because of the potential for media breakdown over time.

All supply air ductwork should be insulated on the exterior, but return air ductwork should be insulated on the exterior if the difference between the duct condition and the ambient space is enough to make condensation possible. All duct joints, both in the duct itself and in the insulation, should be sealed against air leaks. Interior ductwork insulation/lining is not recommended in cultural heritage applications because of the potential for trapping moisture and the eventual breakdown into particulates that can deposit in collections areas. Where external insulation is not possible, or there are significant sound attenuation needs, interior duct lining may be considered if it meets hospital and health care facility standards: fiber free, closed cell, antimicrobial coated, low-VOC certified, and moisture and mold resistant. These sections must be clearly labeled on design drawings and should be discussed with institutional staff regarding ongoing maintenance and eventual replacement.

Particulate filtration is essential for removal of contaminants that could foul the HVAC system, as well as particles that might degrade or deface collections being preserved. For this reason, particulate filtration is addressed here in two steps: prefiltration and fine-particulate filtration.

Physical location of the filtration stages affects HVAC system performance, energy use, and ultimately protection of collections. In upstream filtration, all filtration is placed upstream of the cooling/heating and fans in system. All filters are essentially prefilters, even though fine-particle filtration may be present in the HVAC system. When all filters are upstream, it is common practice to stack filters, with a MERV 7 or 8 prefilter on a MERV 14 or 15 fine particle filter in the same section. This is done to preserve the life of the fine-particle filter. However, evaluate whether the cost of energy outweighs the savings of the fine-filter life: most collection areas are fairly clean environments. Eliminating the prefilter can be a very effective improvement to airflow and energy savings.

In upstream/downstream filtration, prefiltration is installed upstream of the cooling/heating and fan components and final filters downstream of all mechanical equipment. This more effective approach prevents a failure of HVAC components from fouling the ductwork and preservation areas downstream. In this case, it is important to filter at a minimum of MERV 11 for any chilled-water systems to prevent microbial fouling of the coils. MERV 7 or 8 is at best only 50% effective in this particle size category, and does not provide the protection needed against biofilm growth.

Prefiltration is required to prevent fouling in cooling coils and dust build-up in the fan, ductwork, or other HVAC components. It also protects and prolongs the functional service life of gas-phase filters and fine-particle filters. These fouling-size particles and microbes tend to be between 2 and 6 μm and accumulate into biofilms on coils and other wet areas in HVAC systems. MERV 7 or 8 filters by definition will not protect HVAC systems from this particle challenge and should be replaced by a minimum of MERV 11.

In some cases, prefiltration for HVAC system protection is the only filtration in a collection area. If this is the case, MERV 13 is recommended.

Fine-particulate filtration protects artifacts and collections in the facility. These accumulation-size particles fall in the MERV E-1 range of particles (0.3 to 1 μm). Removal efficiencies of a minimum of 85% of the E-1 range are sufficient for preservation of most collections. MERV 15 filters by definition are minimum 85% in the E-1 range, and minimum 90% in both the E-2 (1 to 3 μm) and E-3 (3 to 10 μm) ranges.

Some collections may require efficiencies higher than MERV 15 for long-term preservation. Options include microenclosures with minimal airflow and separate filtration, or HEPA (99.97% at 0.3 μm) filtration for the entire common area. Whenever HEPA filtration is used as the final filtration, seriously consider upgrading the prefiltration to protect the life of the HEPA filters.

Framing systems should be able to seal the air filters without bypass air leakage in housings and unit access doors. System designs with positive locking mechanisms for filters are beneficial, as is using gaskets rather than framing components on filters.

High-voltage electrostatic air cleaners should be used with caution because of their potential to generate ozone, which can damage collections (see Table 9).

Outdoor air infiltration of gaseous pollutants, materials off-gassing in new construction, and similar off-gassing of furnishings may put some collections at risk. Sensitive collections, such as those containing some metals and alloys, film, various papers and low-fired ceramics, should be carefully enclosed or controlled by an active gas-phase filtration system, depending on which method is most appropriate for reaching the desired preservation target (see the section on Control Strategies for Objects with High Vulnerability to Pollutants).

The primary compounds of concern include hydrogen sulfide, nitrogen dioxide, ozone, sulfur dioxide, and undesired volatile organic compounds, all of which are removable with molecular filtration. The specific sorbent must be chosen for the various gaseous contaminants indigenous to the facility, because removal and retention properties are not all the same. Some gases are easily removed with activated carbon, whereas others may require treated sorbents or beds. Using potassium-permanganate-treated media is not recommended because of the risk from the highly oxidative dust to collections and space surfaces.

Service life of molecular filtration media should be carefully evaluated. Much service life testing has focused on the potential weight removal capacity of the sorbent when immersed in a challenge chemical; although this information may be useful, its ability to remove pollutants from an airstream may not be achievable even though the sorbent is not spent.

If a molecular filtration system is used, some gas-filter designs require dusting filters downstream, because the aging gas filters can release dust that can affect the HVAC system and collection.

Gas-phase filtration system design should also consider institutional staffing and preventive maintenance practices; infrastructure must be able to support timely media replacement, in both labor and material cost. Consider gaseous contaminant monitoring and analysis before fully adopting gas-phase filtration; designers may choose to include the gaseous filter section in the design, but wait to install media until need has been determined.

Air volume design can depend on several factors, including

In most cultural heritage applications, variable-air-volume (VAV) systems, in single-zone or multizone configurations, are appropriate.

Note: multizone VAV systems should not be used to serve multiple functional zones. Rather, the approach should be to account for varying loads within the same functional zone. (See the sections on Zoning and on Zoning/Functional Organization)

Multizone VAV has the distinct advantage of providing better temperature and relative humidity control to individual subzones with different sensible loads; both multi- and single-zone VAV design can achieve significant energy-savings in climates with diurnally or seasonally variable sensible and latent loads. Single-zone VAV designs in cultural heritage typically vary air volumes and occasionally sensible reheat temperatures; cooling coils, if used for dehumidification, most often have a single leaving air set point.

Constant-volume system designs may be necessary in certain scenarios. Some geological collections can emit radon, and may require constant volume and increased air change rates. Interior zones (e.g., storage rooms surrounded by a perimeter zone and with limited outdoor air) and some climates may experience essentially the same latent and sensible loads constantly, limiting the effectiveness of VAV designs. Even in these cases, initial designs should consider the inclusion of at least a VFD: actual air volume requirements may be different than design modeling, and the VFD allows adjustment of the optimal volume, even if the system operates at a constant volume moving forward.

For small volumes and applications, such as some cold/frozen storage vaults, constant volume (CV) may be the most practical design, whether based on equipment airflow requirements or simply because a VAV design offers no noticeable preservation or energy advantage.

Institutional capabilities and infrastructure must inform air volume and airflow designs. VAV control and equipment can quickly add to control/operational complexity and maintenance requirements; consider staff capabilities and/or the availability of qualified contractors before finalizing the design.

A reheat system can present problems if improperly applied. In many institutions, terminal reheat with steam or hot-water coils located near or over collection spaces causes chronic problems from steam and water leaks. Subzone humidification control guidance often suggests placing the humidifier downstream from the terminal reheat coil; if the reheat coil is located near or over collection spaces, preventive maintenance on humidifiers further complicates the maintenance requirements. Reheat systems for collections zones are most effective when reheat coils and humidifiers are installed entirely within the mechanical space, instead of at the terminal end, feeding through what is effectively a multizone distribution system.

A multizone air handler with sufficient dehumidification capacity at the primary unit, zone reheat, and zone humidification can be a stable and relatively energy-efficient solution. However, multizone systems without individual zone reheat and zone humidification have proved problematic for many institutions, requiring retrofit of zone equipment for stable humidity control. With proper layout and complementary equipment, a multizone system can reduce the amount of reheat and be very energy efficient. Multizone systems should consider dehumidification in the outdoor airstream or upstream of the heating coil to ensure that sufficient dehumidification is available to the entire airstream. Note that multizone designs are still best applied to spaces with similar dehumidification requirements; mixing collections and noncollections spaces in the same multizone design typically leads to inefficiency and difficulty maintaining space conditions. Future renovation or changes in space usage can significantly impact design viability, again leading to issues with efficiency and management of the preservation environment.

As with multizone systems, dual-duct designs can work well in cultural heritage if zoning is carefully considered. Downstream mixing boxes must be controlled and maintained to guard against overcooling and the resultant high-relative-humidity conditions. A critical consideration for dual-duct designs is a separate dehumidification coil upstream of both the hot and cold decks. This separate cooling coil, distinct from the one in the cold deck, is used during dehumidification demand. Air can be cooled to dew point even if it eventually flows through the hot deck. Without this feature, moist return air or outdoor air could be warmed in the hot deck and delivered back to the room without being dehumidified. An alternative is to locate a single cooling coil upstream of both decks, or in the outdoor air.

Fan-coil units have been problematic when placed in and above collection areas. Fan-coil units expand and decentralize maintenance, requiring maintenance in collection areas and a net increase in overall facility maintenance. Because they cool locally, they need condensate drains, which can leak or back up over time. As all-water systems, they require four pressurized-water pipes to each unit, increasing the chance of piping leaks in collection areas.

Fan-powered mixing boxes are usually inappropriate for cultural heritage facilities. Although fan-powered mixing boxes can help ensure air circulation to suppress mold growth, they do not allow effective air filtration for particles and gases. These fans also increase local maintenance requirements and present an added fire risk. If they include reheat, there is an added risk from leaks (with water or steam reheat) or fire (with electric reheat).