Moisture Dry-Out Capability of Steel-Faced Mineral Wool Insulated Sandwich Panels

sustainability

Article Moisture Dry-Out Capability of Steel-Faced Mineral Wool Insulated Sandwich Panels

Kristo Kalbe 1,*, Hubert Piikov 1 and Targo Kalamees 1,2

1 Nearly Zero Energy Buildings Research Group, Department of Civil Engineering and Architecture, Tallinn University of Technology, 19086 Tallinn, Estonia; [email protected] (H.P.); [email protected] (T.K.) 2 Smart City Centre of Excellence (Finest Twins), Tallinn University of Technology, 19086 Tallinn, Estonia * Correspondence: [email protected]

Received: 25 September 2020; Accepted: 21 October 2020; Published: 30 October 2020

Abstract: Moisture dry-out from steel-faced insulated sandwich panels has previously received little attention from researchers. This paper reports the results from laboratory tests and dynamic heat, air, and moisture transport simulations of the moisture dry-out capabilities of a steel-faced sandwich panel with a mineral wool core. Three test walls (TWs) with dimensions of 1.2 m 0.4 m 0.23 m × × were put above water containers to examine the moisture transport through the TWs. A calibrated simulation model was used to investigate the hygrothermal regime of a sandwich panel wall enclosure with different initial moisture contents and panel joint tightening tapes. The moisture dry-out capacity of the studied sandwich panels is limited (up to 2 g/day through a 30-mm-wide and 3-m-long vertical joint without tapes). When the vertical joint was covered with a vapour-permeable tape, the moisture dry-out was reduced to 1 g/day and when the joint was covered with a vapour-retarding tape, the dry-out was negligible. A very small amount of rain would be enough to raise the moisture content to water vapour saturation levels inside the sandwich wall, had the rain ingressed the enclosure. The calculated time of wetness (TOW) on the internal surface of the outer steel sheet stayed indefinitely at about 5500 h/year when vapour-retarding tapes were used and the initial relative humidity (RH) was over 80%. TOW stabilised to about 2000 h/year when a vapour-permeable tape was used regardless of the initial humidity inside the panel. A vapour-permeable tape allowed moisture dry-out but also vapour diffusion from the outside environment. To minimise the risk of moisture damage, avoiding moisture ingress during construction time or due to accidents is necessary. Additionally, a knowledge-based method is recommended to manage moisture safety during the construction process.

Keywords: moisture safety; sandwich panels; moisture dry-out; laboratory test; HAM modelling

1. Introduction Insulated sandwich panels are layered structures which have two facings and an insulating core. The facings are relatively thin and strong and could be made out of varied materials, e.g., metals (commonly steel) or ﬁbre-reinforced composites and wood-based materials [1,2]. The majority of sandwich panel producers in Europe focus on panels with metal facings and rigid plastic foam or mineral wool for the core material [2]. Plastic foams give the highest values of thermal insulation but panels with mineral wool can also reach adequate thermal insulating properties. Such panels are particularly well suited for walls and roofs [2]. However, plastic foam materials are combustible and mineral wool is preferred when there is a particular requirement for ﬁre safety [2]. Recent research suggests that mineral wool has a lower environmental impact than hydrocarbon-based insulation materials [3,4] and, thus, might be of interest as the more sustainable choice of the two. For a building

Sustainability 2020, 12, 9020; doi:10.3390/su12219020 www.mdpi.com/journal/sustainability Sustainability 2020, 12, 9020 2 of 18 Sustainability 2020, 12, x FOR PEER REVIEW 2 of 18 literatureto be sustainable, regarding it must,durability among and other sustainability aspects, maintain and brought long-term out that durability. one of the Peris most Mora efficient reviewed ways toliterature achieve regarding sustainability durability is to and improve sustainability the durability and brought of the out construction that one of the works most and effi cient that waysgreater to materialachieve sustainabilitydurability leads is to to improve less time the and durability resources of required the construction to maintain works it [ and5]. The that annual greater installed material areadurability of these leads type to of less panels time andhas resourcesreached up required to 130 tomillion maintain m2 itin [ 5Europe]. The annual [6]. Thus, installed the evaluation area of these of thetype hygrothermal of panels has reached performance up to 130 and million from m this2 in Europe the durability [6]. Thus, of the the evaluation panels of could the hygrothermal interest the constructionperformance industry. and from this the durability of the panels could interest the construction industry. MoistureMoisture dry dry-out-out from from structures structures is an important part part of of the the overall overall hygrothermal hygrothermal performance performance ofof the building envelope [7 [7–9].9]. Information Information about about the the dry dry-out-out capability capability is is necessary necessary to to design design and and constructconstruct buildings buildings in in a a manner manner where where there there would would not not be be any any t threatshreats to to human human health health as as a a result result of of highhigh humidity humidity [10]. [10]. In In Sweden, Sweden, an an industry industry standard standard for for including including moisture moisture safety safety in in the the construction construction process is commonly used used [11]. [11]. However, However, research research from from Sweden Sweden shows shows that that,, even even after after the the industry industry standardstandard has has been been i inn use for years, the knowledge in in moisture moisture safety safety is is still low and designers have have problems assess assessinging moisture risks [12]. [12]. Therefore, Therefore, it it is is necessary necessary for for researchers researchers to to provide provide guidelines guidelines forfor stakeholders stakeholders in in the the construction construction business business about about dry dry building building desi designgn and and moisture moisture-safe-safe practices. practices. KnowingKnowing the the critical critical environment environment conditions conditions during during construction construction time time and and the moisture the moisture dry-out dry-out rate ofrate the of usedthe used structures, structures, stakeholders stakeholders responsible responsible for for moisture moisture safety safety can can make make knowledge knowledge-based-based decisions.decisions. Experience Experience shows shows tha thatt general general guidelines guidelines such such as as “ “avoidavoid moisture moisture ingress” ingress” or or “ “protectprotect moisturemoisture-sensitive-sensitive materials materials during during construction construction or or storage” storage” are are not not thoroughly thoroughly followed followed (Figure (Figure 1).1). MoreMore specificspecific guidelines, such such as as in in which which conditions conditions (e.g. (e.g.,, relative relative humidity humidity of of outside outside air, air, rain amount)amount) it it is is safe safe to to install install specific specific details, details, in in which which conditions conditions the the details details must must be be stored, stored, etc etc.,., could could bebe useful. To To give give such such guidelines, guidelines, it it is is necessary necessary to to know know the the moisture dry dry-out-out capability and and the the conditioconditionsns when critical situations occur.

(a) (b)

FigureFigure 1. 1. SandwichSandwich panel panel insulating insulating core core exposed exposed to outside to outside environment environment during during construction construction (a) and (a) storageand storage (b). Photos (b). Photos taken taken by the by authors. the authors.

However,However, there has been little research regarding the moisture dr dry-outy-out of steel steel-faced-faced sandwich panels. Most Most research research regarding regarding the the durability durability of these of these panels panels has focused has focused on the mechanical on the mechanical resistance resistanceand only aand few only acknowledge a few acknowledge the influence the ofinfluence moisture. of moisture. Pfeiffer brings Pfeiffer out brings in his out thesis in his about thesis the aboutdurability the durability of sandwich of sandwich panels that panels “temperature that “temperat and humidityure and ahumidityffect the ageing affect the of mineral ageing woolof mineral cores woolin sandwich cores in panels” sandwich and thatpanels” “it is and expected that “ thatit is increased expected humidity that increased and increased humidity temperature and increased lead temperatureto a faster loss lead in strength”to a faster because loss in “curedstrength” urea-formaldehyde because “cured urea (UF)-formaldehyde resin is not completely (UF) resin stable is not to completelyhydrolysis” stable [13]. to Being hydrolysis” a component [13]. Being of the a mineralcomponent wool of binder,the mineral UF resinwool makesbinder, its UF hydrolysis resin makes to itsinfluence hydrolysis the mechanical to influence properties the mechanical of the mineral properties wool of [13 the]. mineral Moisture-induced wool [13]. damageMoisture to-induced mineral damagewool is alsoto mineral reported wool in ais recentalso reported study where in a recent the authors study determinedwhere the authors that the determined mean compressive that the meanstrength compressive of a mineral strength wool sample of a mineral that had wool a prolonged sample that presence had a of prolonged elevated moisturepresence contentof elevated was moisture93% lower content than the was declared 93% lower value than [14 the]. Pfei declaredffer reports value that [14]. a Pfeiffer combination reports of that elevated a combination temperature of elevated temperature (65 °C) and humidity (≈100% relative humidity) decreases the tensile strength of a sandwich panel to 60% of its initial strength in about a year, stabilising to around 40% to 50% of Sustainability 2020, 12, 9020 3 of 18

(65 C) and humidity ( 100% relative humidity) decreases the tensile strength of a sandwich panel to ◦ ≈ 60% of its initial strength in about a year, stabilising to around 40% to 50% of the initial strength after a few more years. Although Pfeiffer notes that the scatter in results is great and the results depend largely on sample variation, it is evident that high relative humidity (RH) levels are to be avoided in the sandwich panel. In the book Lightweight Sandwich Construction [2], it is also highlighted that high humidity rather than high temperature is the degrading factor for mineral wools in the sandwich panel. The authors of the book also mention that internal moisture variations in the sandwich panel could lead to the corrosion of the steel faces. The back faces are usually coated to inhibit corrosion. The principle of the test method to evaluate corrosion resistance of the coatings is to expose the test panel to continuous water condensation for 500 h or up to 1500 h [15]. However, if there is a water leakage situation, the high humidity conditions could last longer. Increasing moisture content also increases the thermal conductivity of mineral wool [16]. Studies show that significant reduction in the thermal insulation function is already present at relatively low moisture contents of 5–20% by volume of the mineral wool insulation [17]. Laukkarinen et al. made a series of laboratory measurements with steel-faced mineral wool sandwich panels in a water leakage situation [18]. Their paper focused mostly on measurement techniques and methods to detect moisture leakages, but it was evident from the measurements that the outer layer of the mineral wool core experienced condensing conditions repeatedly. The authors reported that the spread of moisture is more influenced by the evaporation speed than the speed of vapour diffusion [18]. Nonetheless, there was no analysis of the moisture dry-out capability of the tested structures. The purpose of this study is to quantify the moisture dry-out capability of steel-faced sandwich panels and investigate the hygrothermal behaviour of the panels when different joint sealing tapes are applied and how the tapes affect the moisture dry-out and the overall hygrothermal performance of steel-faced sandwich panels. The knowledge from the study gives an important input to dry building design and moisture safety management of the construction process. For example, the dry-out rate of a structure is required to evaluate the criticality of wetting incidents or the need for weather protection.

2. Materials and Methods

2.1. Studied Sandwich Panels The studied panels are steel-faced and have a mineral (stone) wool insulation core. The insulation thickness of the panels is 230 mm. The insulation thickness could vary in practice depending on the needed thermal transmittance. The thickness of the external steel sheet is 0.6 mm and of the internal sheet, 0.5 mm. According to the standard installation method, the vertical joints of the panels have a 30 mm wide gap, which must be ﬁlled with insulation and sealed from both sides during the installation of the panels (Figure2a,b). Horizontal joints have elastic sealing strips in place from the factory (Figure2c). Steel faces are vapour tight and only a negligible amount of water vapour is assumed to escape through the horizontal joint, thus, the only area to have an impact on the drying capability of the panels is assumed to be the vertical joint and the tapes covering it. Sustainability 2020, 12, x FOR PEER REVIEW 3 of 18 the initial strength after a few more years. Although Pfeiffer notes that the scatter in results is great and the results depend largely on sample variation, it is evident that high relative humidity (RH) levels are to be avoided in the sandwich panel. In the book Lightweight Sandwich Construction [2], it is also highlighted that high humidity rather than high temperature is the degrading factor for mineral wools in the sandwich panel. The authors of the book also mention that internal moisture variations in the sandwich panel could lead to the corrosion of the steel faces. The back faces are usually coated to inhibit corrosion. The principle of the test method to evaluate corrosion resistance of the coatings is to expose the test panel to continuous water condensation for 500 h or up to 1500 h [15]. However, if there is a water leakage situation, the high humidity conditions could last longer. Increasing moisture content also increases the thermal conductivity of mineral wool [16]. Studies show that significant reduction in the thermal insulation function is already present at relatively low moisture contents of 5%–20% by volume of the mineral wool insulation [17]. Laukkarinen et al. made a series of laboratory measurements with steel-faced mineral wool sandwich panels in a water leakage situation [18]. Their paper focused mostly on measurement techniques and methods to detect moisture leakages, but it was evident from the measurements that the outer layer of the mineral wool core experienced condensing conditions repeatedly. The authors reported that the spread of moisture is more influenced by the evaporation speed than the speed of vapour diffusion [18]. Nonetheless, there was no analysis of the moisture dry-out capability of the tested structures. The purpose of this study is to quantify the moisture dry-out capability of steel-faced sandwich panels and investigate the hygrothermal behaviour of the panels when different joint sealing tapes are applied and how the tapes affect the moisture dry-out and the overall hygrothermal performance of steel-faced sandwich panels. The knowledge from the study gives an important input to dry building design and moisture safety management of the construction process. For example, the dry- out rate of a structure is required to evaluate the criticality of wetting incidents or the need for weather protection.

2. Materials and Methods

SustainabilitySustainability 20202020,, 1122,, xx FORFOR PEERPEER REVIEWREVIEW 44 ofof 1818

22..22.. LaboratoryLaboratory TestTest TheThe laboratorylaboratory testtest (Figure(Figure 3)3) waswas setset upup inin aa controlledcontrolled environmentenvironment toto measuremeasure thethe drydry--outout capabilitycapability andand studystudy thethe effectseffects ofof aa vapourvapour--permeablepermeable andand aa vapourvapour--retardingretarding windwind barrierbarrier tapetape onon thethe dryingdrying outout ofof moisturemoisture fromfrom thethe steelsteel--facedfaced sandwichsandwich panel.panel. ThreeThree segmentssegments (test(test walls,walls, TWs)TWs) (a) (b) (c) werewere extractedextracted fromfrom twotwo largerlarger sandwichsandwich panelspanels ((FigureFigure 44).). TestTest wallwall 11 (TW1)(TW1) andand testtest wallwall 22 (TW2)(TW2) werewereFigure Figure cutcut fromfrom 2. 2. A Athethe mock mock-up samesame-up panel panelof of a a standard standard andand hadhad vertical vertical thethe insulationinsulation panel panel joint joint (stone(stone ( (aa),), thewool) thewool) respective respective fibresfibres parallel paralleldetail detail drawing drawing toto theirtheir ( (blongerlongerb)) and and a sides. asides. TestTest drawingwalldrawingwall 33 (TW3)(TW3) of of the the horizontal horizontalwaswas cutcut out outpanel panel fromfrom joint joint anotheranother ( (cc).). panelpanel withwith thethe stonestone woolwool fibresfibres perpendicularperpendicular toto itsits longer2.2.longer Laboratory sides.sides. Test ItIt waswas necessarynecessary toto createcreate aa gradientgradient ofof waterwater vapourvapour pressurepressure throughthrough thethe insulationinsulation (stone(stone wool)wool)The ofof thethe laboratory sandwichsandwich test panel.panel. (Figure AA container3container) was set filledfilled up in withwith a controlled waterwater waswas environment putput underunder thethe to TWs measureTWs (Figure(Figure the 3b). dry-out3b). TheThe TWscapabilityTWs werewere 10 and10 mmmm study apartapart the fromfrom effects thethe of waterwater a vapour-permeable levellevel andand carecare waswas and takentaken a vapour-retarding notnot toto disturbdisturb thethe wind waterwater barrier levellevel tape whilewhile on weighingtheweighing drying thethe out TWsTWs of moisturetoto avoidavoid watewate fromrr thecontactcontact steel-faced withwith thethe sandwich insulation.insulation. panel. TheThe TWsTWs Three werewere segments preparedprepared (test inin walls,suchsuch aa TWs) wayway thatwerethat waterwater extracted vapourvapour from couldcould two escapeescape larger them sandwichthem onlyonly panels throughthrough (Figure aa designateddesignated4). Test wallsurface,surface, 1 (TW1) whichwhich and waswas test onon wall thethe oppositeopposite 2 (TW2) sidewereside ofof cut thethe from waterwater the container.container. same panel AllAll and otherother had sidessides the insulation ofof thethe TWsTWs (stone werewere wool) sealedsealed fibres vapourvapour parallel tighttight to withwith their aa longer butybutyll band sides.band andTestand aluminiumaluminiumwall 3 (TW3) foilfoilwas covering.covering. cut out ThisThis from setupsetup another imitatedimitated panel thethe with circumstancescircumstances the stone wool thatthat wouldwould fibres arise perpendiculararise ifif waterwater leakedleaked to its intolongerinto thethe sides. sandwichsandwich panelpanel andand accumulatedaccumulated inin itsits bottombottom partpart (e.g.(e.g.,, inin thethe plinthplinth rail).rail).

232232 mmmm ONLYONLY TOP TOP SIDESSIDES OPENOPEN TO TO 400400 mmmm VAPOURVAPOUR DIFFUSIONDIFFUSION

300 mm 200 mm 300 mm 200 mm SENSORSENSOR 33 STUDIEDSTUDIED TAPESTAPES 300300 mmmm SENSORSENSOR 22

300300 mmmm SENSORSENSOR 11

WATERWATER 300300 mmmm UNDERUNDER PANELSPANELS ((aa)) ((bb)) ((cc))

FigureFigure 3. 3. 3D3D model model ofof a a test test wall wall (TW) (TW) ( (aa)) with indicated sensor sensor placement placement (red(red dots) dots) and and photos photos of of thethe TWsTWs (( bb)) inin thethe TalTechTalTech nZEB nZEB Test TestTest FacilityFacility and and of of the the top top sides sides of ofof the thethe TWs TWsTWs with withwith drying drying apertures apertures coveredcovered with with tapes tapes ( (cc).).).

FigureFigure 4.4. LocationLocationLocation ofof thethe testtest wallwall (TW)(TW) cutcut cut-outs--outsouts (blue(blue dasheddashed line)line) inin tt thehehe largerlarger sandwich sandwich panels. panels.

NineNine temperaturetemperature andand RHRH (t&RH)(t&RH) sensorssensors (HOBO(HOBO UX100UX100--023A)023A) werewere installedinstalled onon thethe centrecentre lineline ofof thethe TWsTWs inin thethe middlemiddle ofof thethe insulationinsulation layer.layer. TheThe sensorssensors werewere distributeddistributed atat equalequal distancesdistances (300(300 mm)mm) fromfrom eacheach otherother andand fromfrom thethe wawaterter levellevel andand toptop sideside (Figure(Figure 3a).3a). AllAll thethe penetrationspenetrations duedue toto thethe sensors’sensors’ cablescables werewere sealedsealed withwith aa butylbutyl bandband andand anan aluminiumaluminium tapetape toto minimiseminimise theirtheir influenceinfluence onon the the measurements. measurements. AnotherAnother t&RH t&RH sensor, sensor, which which logged logged the the ambient ambient air air temperaturetemperature andand RHRH duduringring thethe test,test, waswas installedinstalled directlydirectly aboveabove thethe TWs.TWs. TheThe accuracyaccuracy ofof thethe t&RHt&RH sensorssensors waswas validatedvalidated withwith thethe helphelp ofof saturatedsaturated saltsalt solutions,solutions, whichwhich produceproduce aa fixedfixed humidityhumidity level,level, asas describeddescribed byby GreenspanGreenspan [19].[19]. TheThe sensorssensors werewere placedplaced sequentiallysequentially Sustainability 2020, 12, 9020 5 of 18

It was necessary to create a gradient of water vapour pressure through the insulation (stone wool) of the sandwich panel. A container filled with water was put under the TWs (Figure3b). The TWs were 10 mm apart from the water level and care was taken not to disturb the water level while weighing the TWs to avoid water contact with the insulation. The TWs were prepared in such a way that water vapour could escape them only through a designated surface, which was on the opposite side of the Sustainability 2020, 12, x FOR PEER REVIEW 5 of 18 water container. All other sides of the TWs were sealed vapour tight with a butyl band and aluminium foil covering.above the Thisaqueous setup solutions imitated of MgCl, the circumstances NaCl and K2SO that4, which would produced arise if wateran environment leaked into with the an sandwich RH panelof and 32.78% accumulated ± 0.16, 75.29% in its± 0.12 bottom and 93.58% part (e.g., ± 0.55, in respectively, the plinth rail). at 25 °C. The sensors were held above theNine saturated temperature salt solutions and RH for at (t&RH) least 12 sensorsh after the (HOBO readings UX100-023A) stabilised. The werecorrection installed values on reached the centre line ofup the to +3.7% TWs inat the79% middle RH but of stayed the insulation below +2.5% layer. at 33% The and sensors 98% RH, were being distributed below +1.5% at equal for most distances (300 mm)pointsfrom at these each RH other levels and(Figure from 5). theThis waterwas considered level and when top reporting side (Figure the 3resultsa). All of thethis penetrationspaper. due to the sensors’ cables were sealed with a butyl band and an aluminium tape to minimise their influence on the measurements.4.0 Another t&RH sensor, which logged the ambient air temperature and 3.5 RH during the test, was installed3.0 directly above the TWs. The accuracy of the t&RH sensors2.5 was validated with the help of saturated salt solutions, which Sustainability 2020, 12, x FOR PEER REVIEW 5 of 18 produce a fixed humidity level,2.0 as described by Greenspan [19]. The sensors were placed sequentially 1.5 above the aqueous solutions of MgCl, NaCl and K 2 SO44, which produced an environment with an RH above the aqueous solutions of1.0 MgCl, NaCl and K SO , which produced an environment with an RH of 32.78%of 32.78%0.16, ± 0.16, 75.29% 75.29% 0.12± 0.12 and and 93.58% 93.58% ± 0.55,0.55, respectively, respectively, at at25 25°C.◦ C.The The sensors sensors were were held heldabove above ± ± 0.5 ± the saturatedthe saturated salt salt solutions solutions t&RH the for Correction for for at0.0 at least least 12 12 h h afterafter the readings readings stabilised. stabilised. The The correction correction values values reached reached up toup+ 3.7%to +3.7% at 79%at 79% RH RH but but humidity relative sensor value stayed stayed20 below below30 + 40+2.5%2.5%50 at 33% 33%60 and and70 98% 98%80 RH, RH,90 being being100 below below +1.5%+1.5% for most for most Relative humidity, % R1 R2 R3 R4 R5 pointspoints at these at these RH RH levels levels (Figure (Figure5). 5). This This was was considered consideredR6 when R7 reporting reportingR8 R9 the the results results of this of this paper. paper.

Figure 5. Correction values4.0 for the temperature and relative humidity (t&RH) sensors (R1–R9) at different RH values. 3.5 3.0 The TWs were kept in isothermal2.5 conditions, i.e., the ambient temperature and RH were the same around all sides of the TWs2.0 (Figure 6, water vapour pressure 793–1266 Pa), with the exception of the bottom side of the TWs 1.5where the RH was ≈ 100% due to the water vapour pressure (2642–3166 Pa) generated from the water1.0 in the container beneath. This setup created a sufficient water vapour gradient for the experiment. 0.5 Correction for the t&RH t&RH the for Correction 0.0

The designated drying humidity relative sensor value surface20 area30 of40 the TWs50 was60 changed70 80 throughout90 100 the test, being either a fully open top side (230 mm × 400Relative mm) orhumidity, an aperture % R1of 30 mmR2 × R3375 mm.R4 TheR5 width of the aperture R6 R7 R8 R9 (30 mm) was chosen to reflect the width of the standard vertical joint for these types of sandwich panels. For the purpose of measuring the water vapour saturation speed, there was also a test FigureFigure 5. Correction 5. Correction values values for for the the temperature temperature and relative relative humidity humidity (t&RH (t&RH)) sensors sensors (R1– (R1–R9)R9) at at sequence when TW1 had all sides sealed. Figure 7 shows a timeline describing the series of diﬀerentdifferent RH values.RH values. measurements and changes made to the dry-out aperture and the environment surrounding the TWs throughoutThe TWsThe TWs were the were test. kept Akept in vapour isothermal in isothermal-permeable conditions, conditions, (with an i.e., S di.e. or the, equivalentthe ambient ambient air temperature temperature layer thickness and and of RH RH 0.05were were m) tape thethe same aroundandsame all a aroundvapour sidesof -allretarding the sides TWs of (Sthe (Figured = TWs 15 m) 6(Figure, watertape were6, vapour water used vapour pressureon the pr apertureessure 793–1266 793 to– study1266 Pa), Pa), withthe witheffect the the exceptionof exceptionthe tapes of the (Figure 3c). bottomof the side bottom of the side TWs of the where TWs thewhere RH the was RH was100% ≈ 100% due due to the to the water water vapour vapour pressure pressure (2642 (2642–3166–3166 Pa) In addition to the data gathered with≈ t&RH sensors, all the TWs were regularly weighed to generatedPa) generated from the from water the water in the in containerthe container beneath. beneath. ThisThis setup created created a sufficient a suﬃcient water water vapo vapourur determine the mass of dried-out (evaporated) water. A calibrated Kern DS 30K0.1L platform scale gradientgradient for thefor experiment.the experiment. was usedThe designated with a maximum drying weighing surface area capacity of the o TWsf 30 kg,was readability changed throughout of 0.1 g and the repeatability test, being ofeither 0.2 g. a

fully open top side (230 mm40 × 400 mm) or an aperture of 30 mm × 375 mm.100 The width of the aperture (30 mm) was chosen to reflect35 the width of the standard vertical joint for90 these types of sandwich panels. For the purpose 30 of measuring the water vapour saturation speed,80 there was also a test

C 25 70

sequence when TW1 had20 all sides sealed. Figure 7 shows a timeline60 describing the series of measurements and changes15 made to the22 dry - 23 -Cout aperture31 and - 35 Cthe Outdoorenvironment50 surrounding the TWs 10 40 d throughout the test. A vapour5 -permeable (with an S or equivalent air layer30 thickness of 0.05 m) tape

and a vapour-retarding (Sd0 = 15 m) tape were used on the aperture to study20 the effect of the tapes Temperature, Temperature, (Figure 3c). -5 10 % Relativehumidity, -10 0 In addition to the data0 gathered7 14 21with28 t&RH35 42 sensors,49 56 63all 70the 77TWs84 were regularly weighed to

determine the mass of dried-out (evaporated) water.Temperature A calibrated around test Kernwalls DS 30K0.1L platform scale Time, days was used with a maximum weighing capacity of 30Relative kg, readability humidity around of test 0.1 walls g and repeatability of 0.2 g.

FigureFigure40 6. 6.Ambient Ambient conditionsconditions during during the the laboratory laboratory test.100 test. 35 90 30 80

C 25 70

20 60 15 22 - 23 C 31 - 35 C Outdoor 50 10 40 5 30

0 20 Temperature, Temperature, -5 10 % Relativehumidity, -10 0 0 7 14 21 28 35 42 49 56 63 70 77 84

Temperature around test walls Time, days Relative humidity around test walls Figure 6. Ambient conditions during the laboratory test. Sustainability 2020, 12, 9020 6 of 18

The designated drying surface area of the TWs was changed throughout the test, being either a fully open top side (230 mm 400 mm) or an aperture of 30 mm 375 mm. The width of the aperture × × (30 mm) was chosen to reﬂect the width of the standard vertical joint for these types of sandwich panels. For the purpose of measuring the water vapour saturation speed, there was also a test sequence when TW1 had all sides sealed. Figure7 shows a timeline describing the series of measurements and changes made to the dry-out aperture and the environment surrounding the TWs throughout the test. A vapour-permeable (with an Sd or equivalent air layer thickness of 0.05 m) tape and a vapour-retardingSustainability 2020, 12 (S, xd FOR= 15 PEER m) tapeREVIEW were used on the aperture to study the eﬀect of the tapes (Figure6 3ofc). 18

Water under TWs Measuring moisture dry-out

Days: 0 26 32 42 53 70 84 88 Vapour-permeable Fully closed top surface TW1 tape (Sd = 0.05 m) Tapes removed from aperture to measure saturation speed over aperture Vapour-retarding tape Fully open top surface to measure TW2 (Sd = 15 m) Tapes removed from aperture water vapour permeability of mineral wool over aperture Fully open top surface to measure Vapour-permeable TW3 water vapour permeability of tape (Sd = 0.05 m) Tapes removed from aperture mineral wool over aperture

tambient ≈ 22-24 C tambient ≈ 31-35 C tambient ≈ 0-10 C RHambient ≈ 31-34% RHambient ≈ 25% RHambient ≈ 90-100% No dry-out Dry-out through fully open top surface

Dry-out only through 30 mm x 375 mm aperture (with or without tapes) Studying temperature variation effects

FigureFigure 7.7. TimelineTimeline (in (in days) days) of of the the laboratory laboratory test test with with descriptions descriptions of various of various parts parts of the of test. the Each test. Eachhorizontal horizontal band band describes describes one oneTW. TW. Distinct Distinct colours colours indicate indicate changes changes to the to the top top surface surface of ofthe the TWs. TWs.

2.3. HAMIn addition Modelling tothe data gathered with t&RH sensors, all the TWs were regularly weighed to determine the mass of dried-out (evaporated) water. A calibrated Kern DS 30K0.1L platform scale was used withThe numerical a maximum simulation weighing tool capacity Delphin of 30 5 kg, was readability used for of the 0.1 combined g and repeatability heat, air and of 0.2 moisture g. (HAM) transport modelling. Delphin is a well-developed, advanced and validated software suitable 2.3.for building HAM Modelling sciences [20,21]. The simulation models were calibrated against the measurements with sufficient accuracy as The numerical simulation tool Delphin 5 was used for the combined heat, air and moisture (HAM) described by Kalbe et al. [22]. A 2D model was necessary to achieve a good agreement between the transport modelling. Delphin is a well-developed, advanced and validated software suitable for measured data and simulation results. Kalbe et al. [22] reported a good agreement of the measured building sciences [20,21]. and simulated drying processes, but there was a possibility of a slight overestimation of drying. The The simulation models were calibrated against the measurements with suﬃcient accuracy as authors compared the measured and simulated RH and water vapour pressures (Figure 8), which described by Kalbe et al. [22]. A 2D model was necessary to achieve a good agreement between the showed sufficiently good agreement, although the simulation tended to underestimate RH in the measured data and simulation results. Kalbe et al. [22] reported a good agreement of the measured and high humidity region. This could have been partly due to the positive shift of registered RH values simulated drying processes, but there was a possibility of a slight overestimation of drying. The authors (up to +3.7% at 79% RH) detected with validating the readings of t&RH sensors over saturated salt compared the measured and simulated RH and water vapour pressures (Figure8), which showed solutions. These discrepancies have been considered when reporting the results of this paper. suﬃciently good agreement, although the simulation tended to underestimate RH in the high humidity region.100 This could have been partly due to the positive shift5000 of registered RH values (up to +3.7% at 22-23 C 31-35 C Outdoor 79% RH)90 detected with validating the readings of t&RH4500 sensors over saturated salt solutions. These discrepancies80 have been considered when reporting the results4000 of this paper. 70 3500 60 3000 50 2500 40 2000 30 Fully open Sd 30-mm-wide 1500 surface 0.05m open surface Sd 20 1000

Relative humidity, % % Relativehumidity, tape Fully open 0.05m 30-mm-wide 10 22-23 C 31-35 C Outdoor 500 surface tape open surface

0 vapour Pa pressure, Water 0 0 7 14 21 28 35 42 49 56 63 70 77 84 0 7 14 21 28 35 42 49 56 63 70 77 84

Time, days Measured_1 Measured_2 Measured_3 Time, days Measured_1 Measured_2 Measured_3 Simulated_1 Simulated_2 Simulated_3 Simulated_1 Simulated_2 Simulated_3 (a) (b)

Figure 8. Comparison of measured and simulated RH (a) and water vapour pressure (b) of TW3. Adapted from Kalbe et al. [22].

Sustainability 2020, 12, x FOR PEER REVIEW 6 of 18

Water under TWs Measuring moisture dry-out

tambient ≈ 22-24 C tambient ≈ 31-35 C tambient ≈ 0-10 C RHambient ≈ 31-34% RHambient ≈ 25% RHambient ≈ 90-100% No dry-out Dry-out through fully open top surface

Dry-out only through 30 mm x 375 mm aperture (with or without tapes) Studying temperature variation effects

Figure 7. Timeline (in days) of the laboratory test with descriptions of various parts of the test. Each horizontal band describes one TW. Distinct colours indicate changes to the top surface of the TWs.

2.3. HAM Modelling The numerical simulation tool Delphin 5 was used for the combined heat, air and moisture (HAM) transport modelling. Delphin is a well-developed, advanced and validated software suitable for building sciences [20,21]. The simulation models were calibrated against the measurements with sufficient accuracy as described by Kalbe et al. [22]. A 2D model was necessary to achieve a good agreement between the measured data and simulation results. Kalbe et al. [22] reported a good agreement of the measured and simulated drying processes, but there was a possibility of a slight overestimation of drying. The authors compared the measured and simulated RH and water vapour pressures (Figure 8), which showed sufficiently good agreement, although the simulation tended to underestimate RH in the high humidity region. This could have been partly due to the positive shift of registered RH values

(upSustainability to +3.7%2020 at, 1279%, 9020 RH) detected with validating the readings of t&RH sensors over saturated7 salt of 18 solutions. These discrepancies have been considered when reporting the results of this paper.

100 5000 90 4500 22-23 C 31-35 C Outdoor 80 4000 70 3500 60 3000 50 2500 40 2000 30 Fully open Sd 30-mm-wide 1500 surface 0.05m open surface Sd 20 1000

Relative humidity, % % Relativehumidity, tape Fully open 0.05m 30-mm-wide 10 22-23 C 31-35 C Outdoor 500 surface tape open surface

0 vapour Pa pressure, Water 0 0 7 14 21 28 35 42 49 56 63 70 77 84 0 7 14 21 28 35 42 49 56 63 70 77 84

Time, days Measured_1 Measured_2 Measured_3 Time, days Measured_1 Measured_2 Measured_3 Sustainability 2020, 12, x FORSimulated_1 PEER REVIEWSimulated_2 Simulated_3 Simulated_1 Simulated_2 Simulated_37 of 18 (a) (b) 2.3.1. Material Properties FigureFigure 8. 8. ComparisonComparison of of measured measured and and simulated simulated RH RH (a (a) )and and water water vapour vapour pressure pressure (b (b) )of of TW3. TW3. AdaptedTableAdapted 1 fromand from FigureKalbe Kalbe et et 9 al. al.present [22]. [22]. the material properties used in the HAM models. The material properties are based on the data provided by the producer of the sandwich panels and Delphin’s 2.3.1. Material Properties default functions (for the properties that depend on the hygric environment) [22]. TheTable steel1 and sheets Figure were9 present omitted the materialfrom the properties simulation used models in the and HAM no models. vapour exchange The material was modelledproperties on are the based corresponding on the data boundary. provided by the producer of the sandwich panels and Delphin’s default functions (for the properties that depend on the hygric environment) [22]. Table 1. Material properties of the stone wool insulation of the sandwich panels [22]. Table 1. Material properties of the stone wool insulation of the sandwich panels [22]. Material Property Value BulkMaterial density Property ρ, kg/m3 Value85 ThermalBulk density conductivityρ, kg/m3 λ, W/(m·K) 850.04 SpecificThermal heat conductivity capacityλ c,W, J/(kg·K)/(m K) 0.04840 · Specific heat capacity c,J/(kg K) 840 Water vapour diffusion resistance· factor μ, - 1.0 Water vapour diffusion resistance factor µ,- 1.0 3 3 EffectiveEffective saturationsaturationθ ,θ m, m3/m/m3 0.90.9 PorosityPorosity θ,, mm33//mm33 0.970.97 1 WaterWater uptake coe coefficientfficient A wA,w kg, kg/(m·s½)/(m s ) 0 0 · 2 Liquid water conductivity k , kg/(m s Pa) 0 Liquid water conductivityl kl, kg/(m·s·Pa)· · 0

900 600 300 0 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0 10 20 30 40 50 60 70 80 90 100 Relative humidity, % Sorption isotherm FigureFigure 9. 9. SorptionSorption isotherm isotherm used used for for the the stone stone wool wool in in the the simulations. simulations.

2.3.2. TheClimate steel Data sheets and were Boundary omitted Conditions from the simulation models and no vapour exchange was modelled on the corresponding boundary. The HAM transport simulations were made with the outside temperature and RH data of the Estonian2.3.2. Climate moisture Data reference and Boundary year (MRY) Conditions for water vapour condensation analysis. The climate data (Figure 10a) correspond to a year where the potential of an assembly to dry out is low and which The HAM transport simulations were made with the outside temperature and RH data of the would occur once in every ten years in Estonia [23]. A simplified relationship of outdoor and indoor Estonian moisture reference year (MRY) for water vapour condensation analysis. The climate data temperatures was used for the indoor boundary (Figure 10b). This corresponds to typical indoor (Figure 10a) correspond to a year where the potential of an assembly to dry out is low and which conditions of heated warehouses and is based on the approach described in EN 15026:2007 [24]. No would occur once in every ten years in Estonia [23]. A simpliﬁed relationship of outdoor and indoor vapour exchange was assumed between the panel and the indoor environment. temperatures was used for the indoor boundary (Figure 10b). This corresponds to typical indoor 100 100 28 90 90 Indoor temperature

80 C 26 80 o 70 70 60 24 60 50

40 C 50 22 30 40 20 20 30 10 0

Relative % Relative humidty, 20 -10 18

10 -20 Indoor temperature,

0 -30 Temperature, 16 06.1995 09.1995 12.1995 03.1996 05.1996 -25 -20 -15 -10 -5 0 5 10 15 20 25 Date, mm.yyyy Relative humidity,% Outside temperature, °C Outdoor temperature, oC (a) (b)

Figure 10. Temperature and RH values of the climate data used in heat, air and moisture (HAM) simulations for the outside (a) and inside (b) boundaries. Sustainability 2020, 12, x FOR PEER REVIEW 7 of 18

2.3.1. Material Properties Table 1 and Figure 9 present the material properties used in the HAM models. The material properties are based on the data provided by the producer of the sandwich panels and Delphin’s default functions (for the properties that depend on the hygric environment) [22]. The steel sheets were omitted from the simulation models and no vapour exchange was modelled on the corresponding boundary.

Table 1. Material properties of the stone wool insulation of the sandwich panels [22].

Material Property Value Bulk density ρ, kg/m3 85 Thermal conductivity λ, W/(m·K) 0.04 Specific heat capacity c, J/(kg·K) 840 Water vapour diffusion resistance factor μ, - 1.0 Effective saturation θ, m3/m3 0.9 Porosity θ, m3/m3 0.97 Water uptake coefficient Aw, kg/(m·s½) 0 Liquid water conductivity kl, kg/(m·s·Pa) 0

900 600 300 0 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0 10 20 30 40 50 60 70 80 90 100 Relative humidity, % Sorption isotherm Figure 9. Sorption isotherm used for the stone wool in the simulations.

2.3.2. Climate Data and Boundary Conditions The HAM transport simulations were made with the outside temperature and RH data of the Estonian moisture reference year (MRY) for water vapour condensation analysis. The climate data (FigureSustainability 10a)2020 correspond, 12, 9020 to a year where the potential of an assembly to dry out is low and which8 of 18 would occur once in every ten years in Estonia [23]. A simplified relationship of outdoor and indoor temperatures was used for the indoor boundary (Figure 10b). This corresponds to typical indoor conditions of heated warehouses and is based on the approach described in EN 15026:2007 [24]. conditions of heated warehouses and is based on the approach described in EN 15026:2007 [24]. No No vapour exchange was assumed between the panel and the indoor environment. vapour exchange was assumed between the panel and the indoor environment.

100 100 28 90 90 Indoor temperature

80 C 26 80 o 70 70 60 24 60 50

40 C 50 22 30 40 20 20 30 10 0

Relative % Relative humidty, 20 18 Sustainability 2020, 12, x FOR PEER REVIEW -10 8 of 18 10 -20 Indoor temperature,

0 -30 Temperature, 16 06.1995 09.1995 12.1995 03.1996 05.1996 2.3.3. Simulation Model -25 -20 -15 -10 -5 0 5 10 15 20 25 Date, mm.yyyy Relative humidity,% Outside temperature, °C Outdoor temperature, oC The HAM simulation model(a) comprised only the stone wool part, with(b ) steel modelled as a boundary without vapour exchange. A model of half of a 6-m-long and 230-mm-thick sandwich panel was Figuremade. 10. The TemperatureTemperature length of 6 andm and is RH RHtypical values values for of of sandwich the the climate climate panels. data data used used The in in30 heat, heat,-mm air- airwide and and jointmoisture moisture (see (Section (HAM)HAM) 2.1) wheresimulationssimulations vapour exchange for the outside could ( a )occur and inside was ( alsob) boundaries. included and the corresponding boundary layer was 2.3.3.simulated Simulation with an Model Sd of 0.05 m, which is the value for a vapour-permeable wind barrier tape. To simplify the model and reduce simulation time, the principle of symmetry was applied. Thus, the finalThe 2D HAM geometry simulation in the modelsimulation comprised model was only 3015 the stonemm by wool 230 mm part, (Figure with steel11). modelled as a boundaryIn a pr withoutevious study, vapour where exchange. the authors A model used of half constant of a 6-m-long climate conditions and 230-mm-thick for the simulations sandwich panel [22], wasit was made. reported The that length a high of 6 RH m is occurs typical in for the sandwich external part panels. of the The panel. 30-mm-wide Thus, the joint RH (seeoutputs Section (P1, 2.1P2,) whereP3) in the vapour current exchange simulations could were occur assigned was also to included 3 mm × 3 and mm the segments corresponding in the outermost boundary layer layer of was the simulatedinsulation with(Figure an 11). Sd of The 0.05 output m, which P1 was is the assigned value for to athe vapour-permeable middle of the diffusion wind barrier open joint, tape. P2 was assignedTo simplify to a point the at model one-quarter and reduce of the simulation panel length time, and the P3 principle was assigned of symmetry to the middle was applied. of the panel Thus, the(far ﬁnalright 2D edge geometry in the simulation in the simulation model geometry). model was 3015 mm by 230 mm (Figure 11).

Figure 11. Geometry of the simulated model (yellow) in the context of a standard vertical joint (on the right). Locations Locations of of simulated simulated RH RH points points are are shown shown with with leaders leaders (P1 (P1 = at= theat the diffusion diffusion open open joint; joint; P2 P2= at= oneat one-quarter-quarter of the of thepanel; panel; P3 =P3 at =halfat halfof the of panel). the panel). A zoomed A zoomed-in-in view of view the ofleft the side left ofside the model of the modelat the diffusion at the diff usionopen joint open is joint given is given on the on bottom the bottom left of left the of figure. the figure. MRY MRY—moisture—moisture reference reference year. year.

InThe a previous simulations study, were where made the with authors two used initial constant moisture climate contents: conditions 0.134 for kg/m the simulations3 and 0.291 [ kg/m22], it3 waswhich reported respectively that a highcorrespond RH occurs to 80% in the RH external and water part ofvapour the panel. saturated Thus, conditions the RH outputs (97% (P1,RH P2,on the P3) insorption the current curve). simulations The latter were (overhygroscopic assigned to 3 moisture) mm 3 mm would segments occur whenin the aboutoutermost 67 g layer of water of the is × insulationsealed into (Figure a sandwich 11). Thepanel output with P1materia was assignedl properties to thedescribed middle earlier of the and diﬀusion with dimensions open joint,P2 of was1 m assigned× 1 m × 0.23 to am point (thus at, with one-quarter a volume of theof 0.23 panel m3 length, 0.23 m and3 × P30.291 was kg/m assigned3 ≈ 6.69 to × the 10− middle2 kg moisture). of the panel The (farsituation right edgewould in be the similar simulation to a test model where geometry). a mineral wool sample is put above free water (without direct contact and capillary water flow) in a sealed box and the mineral wool acquires the maximum amount of water molecules due to adsorption. One hundred per cent RH would mean that the mineral wool is saturated with free water (moisture content ≈ 900 kg/m3; Figure 9). The 80% RH level describes a situation where no water leak has occurred, but the panel has equalised with a surrounding autumn outside environment (for comparison, the 28-day moving average of RH values during the autumn and winter of the selected MRY is constantly over 80%). Thus, for the panel with a volume of 0.23 m3, there would be 0.23 m3 × 0.134 kg/m3 ≈ 3.08 × 10−2 kg of moisture (water vapour). The difference of moisture contents between the 80% RH level and water vapour saturated conditions is 0.157 kg/m3 or 3.61 × 10−2 kg for a panel with the volume of 0.23 m3. The simulations were made with the wall facing north, west and south, considering the solar diffuse and direct radiation data of the MRY climate data. The external surface of the sandwich panel was assumed to be medium grey with an absorption coefficient of 0.6. Additionally, a simulation without any radiation influence was made. All the simulations were started with the beginning of January and the yearly data were treated as cyclic, i.e., data wrapped after 365 days. Sustainability 2020, 12, 9020 9 of 18

The simulations were made with two initial moisture contents: 0.134 kg/m3 and 0.291 kg/m3 which respectively correspond to 80% RH and water vapour saturated conditions (97% RH on the sorption curve). The latter (overhygroscopic moisture) would occur when about 67 g of water is sealed into a sandwich panel with material properties described earlier and with dimensions of 1 m 1 m 0.23 m × × (thus, with a volume of 0.23 m3, 0.23 m3 0.291 kg/m3 6.69 10 2 kg moisture). The situation would × ≈ × − be similar to a test where a mineral wool sample is put above free water (without direct contact and capillary water flow) in a sealed box and the mineral wool acquires the maximum amount of water molecules due to adsorption. One hundred per cent RH would mean that the mineral wool is saturated with free water (moisture content 900 kg/m3; Figure9). ≈ The 80% RH level describes a situation where no water leak has occurred, but the panel has equalised with a surrounding autumn outside environment (for comparison, the 28-day moving average of RH values during the autumn and winter of the selected MRY is constantly over 80%). Thus, for the panel with a volume of 0.23 m3, there would be 0.23 m3 0.134 kg/m3 3.08 10 2 kg of × ≈ × − moisture (water vapour). The difference of moisture contents between the 80% RH level and water vapour saturated conditions is 0.157 kg/m3 or 3.61 10 2 kg for a panel with the volume of 0.23 m3. × − The simulations were made with the wall facing north, west and south, considering the solar diffuse and direct radiation data of the MRY climate data. The external surface of the sandwich panel was assumed to be medium grey with an absorption coefficient of 0.6. Additionally, a simulation without any radiation influence was made. All the simulations were started with the beginning of January and the yearly data were treated as cyclic, i.e., data wrapped after 365 days.

2.3.4. Time of Wetness Time of wetness (TOW) is widely used regarding metallic materials and is defined in the ISO standard 9223:2012 as a “period when the metallic surface is covered by adsorptive and/or liquid films of electrolyte to be capable of causing atmospheric corrosion” [25]. The standard also describes that TOW is calculated as the length of time (in hours per year) during which the relative humidity is greater than 80% at a temperature greater than 0 ◦C and notes that “wetting of surfaces is caused by many factors, for example, dew, rainfall, melting snow and a high humidity level”. TOW categories (Table2) range from internal microclimates (level T 1) to damp climates and unventilated sheds in humid conditions (level T5). A higher TOW indicates higher corrosivity of the atmosphere. The latter is also influenced by airborne salinity (represented by chloride) and the concentration of pollutants (sulphur-containing substances represented by sulphur dioxide) in the atmosphere; but in the context of this study, it is assumed that the concentration of these inside the sandwich panel is at the lowest level.

Table 2. Time of wetness (TOW) and corrosivity categories [25].

TOW, h/Year TOW Level Corrosivity Category for Steel *

10 T1 C1 10 < TOW 250 T C ≤ 2 1 250 < TOW 2500 T C ≤ 3 2–3 2500 < TOW 5500 T C ≤ 4 3 5500 < TOW T5 C3–4 * Assuming the lowest chloride and sulphur dioxide concentration.

A recent study described that when estimating TOW from measurement data, a more accurate approach would be to use values obtained from the surface of the metal material, instead of ambient air RH [26]. The same approach has applied in this study to calculate the TOW of the steel surface inside the sandwich panel. The temperature and RH data used in the TOW calculation is simulated in the point P3 (Figure 11), i.e., in the outermost 3 mm thick layer of the mineral wool core in the middle of the panel. Sustainability 2020, 12, 9020 10 of 18

Sustainability 2020, 12, x FOR PEER REVIEW 10 of 18 3. Results about 1 g/(m2∙h) (Figure 12d,f). This could be considered the maximum dry-out rate under these 3.1. Laboratory Test Results climate conditions. In a later phase of the experiment, when the ambient temperature was increased to 31The °C and first then setof further analyses to 35 examined °C, the dry the-out saturation rate per speed square of metre water of vapour the drying (TW1, surface Figure increased, 12a) and thewhich distribution could be explained of water vapour by the insidegreater the evaporation sandwich energy panel (TW2, due to Figure higher 12 temperatures.c and TW3, Figure Howev 12er,e). Thenthe dry the-out configuration rate remained of modest the drying at best surface being was 1.5– changed1.75 g/(m (see2∙h). Figure It is to7 bein expected the Methods that sectionwith lower for aambient more detailed temperatures, description): the dry only-out rate an aperture is lower of. Unfortunately, 30 mm 375 mmdue wasto overlapping left open to experiments, diffusion and it × coveredwas not with possible a vapour-permeable to measure the drytape-out on TW1 rate when and TW3 the and TWs with were a vapour-retarding placed in outdoor tape conditions. on TW2. AfterNevertheless, the RH readingsthe measured stabilised, RH levels the tapes indicate were that removed, due to leavingthe high a RH 30 mm in the wide outside surface environment, uncovered. Inthe parallel, insulation the dried-out core in the moisture TWs quickly for each saturated past period with for every water TW vapour was determined and in such (TW1, conditions Figure 12 theb; TW2,dry-out Figure capacity 12d andwas TW3,negligible. Figure 12f).

100 3 90 ·h) Fully closed Sd 0.05m 30-mm-wide 2 2.5 80 surface tape open surface 70 2 60 50 1.5 Fully closed S 0.05m 30-mm-wide 40 d 22-23 C 31-35 C Outdoor outrate, g/(m 1 surface tape open surface - 30

20 0.5 Relative humidity, % % Relativehumidity, 10 22-23 C 31-35 C Outdoor 0 0 0 7 14 21 28 35 42 49 56 63 70 77 84

0 7 14 21 28 35 42 49 56 63 70 77 84 Moisturedry

Time, days Time, days TW1, average dry-out rate TW1_sensor 1 TW1_sensor 2 TW1_sensor 3 (a) (b) 100 3 90 Fully open Sd 15m 30-mm-wide

·h) 2.5 80 2 surface tape open surface 70 2 60 g/(m 50 1.5 22-23 C 31-35 C Outdoor 40

Fully open Sd 15m 30-mm-wide 1 outrate, 30 surface tape open surface - 20

0.5 Relative humidity, % % Relativehumidity, 10 22-23 C 31-35 C Outdoor 0 0

0 7 14 21 28 35 42 49 56 63 70 77 84 0 7 14 21 28 35 42 49 56 63 70 77 84 Moisturedry Time, days Time, days TW2, average dry-out rate TW2_sensor 1 TW2_sensor 2 TW2_sensor 3 (c) (d) 100 3

90 ·h)

2 Fully open S 0.05m 30-mm-wide 2.5 d 80 surface tape open surface 70 2 60 50 1.5 22-23 C 31-35 C Outdoor

40 g/(moutrate, Fully open S 0.05m 30-mm-wide - 1 30 d surface tape open surface 20

0.5 Relative humidity, % % Relativehumidity, 10 22-23 C 31-35 C Outdoor 0 0

0 7 14 21 28 35 42 49 56 63 70 77 84 Moisturedry 0 7 14 21 28 35 42 49 56 63 70 77 84

Time, days Time, days TW3, average dry-out rate TW3_sensor 1 TW3_sensor 2 TW3_sensor 3 (e) (f)

Figure 12.12. Measured RH inside TW1 ( a), TW2 ( (c) and TW3 TW3 ( (e)) at at different diﬀerent heights (sensor 1 is located closest to to the the water water container, container, sensor sensor 2 2in in the the middle middle of ofthe the TW TW and and sensor sensor 3 at 3the at highest the highest point) point) and andmeasured measured moisture moisture dry-out dry-out rates ratesof the of TWs the TWs(b,d,f (b) ,duringd,f) during various various parts parts of the of test. the test. Sustainability 2020, 12, 9020 11 of 18

From the data in Figure 12a, it can be seen that the RH levels in the sandwich panel rise quickly over 90% if the panel is sealed vapour tight. Water vapour saturation was reached in every measured point in 26 days. The moisture dry-out rate (Figure 12b) for the corresponding period was negligible with the exception of the measurement results in the ﬁrst week of the test. Some vapour leakages were determined around the water container; these were sealed (Figure 13) and afterwards the dry-out rate was close to 0 g/(m2 h). The same additional sealing was also applied around the water containers of · TW2Sustainability and TW3. 2020, 12, x FOR PEER REVIEW 11 of 18

(a) (b)

FigureFigure 13. 13.During During the measurementthe measurement of water of vapourwater saturationvapour saturation speed in TW1,speed an in unexpected TW1, an reductionunexpected of weightreduction occurred. of weight After occurred. examining After the examining TW, a leakage the point TW, wasa leakage found point (a). Additional was found aluminium (a). Additional foil tapealuminium was added foil (b tape) to prevent was added unwanted (b) to prevent vapour unwanted diﬀusion. vapour diffusion.

TheThe RH dry levels‐out in capacity TW2 and was TW3 close were to distributed0 g/(m2∙h) with equally, the correspondingvapour‐retarding to the (Sd distance 15 m) tape from (Figure the water12d) level. and Thearound water 0.5 vapour g/(m2∙h) was with able the to dryvapour out‐ withoutpermeable additional (Sd 0.05 resistance m) tape (Figure (up to day 12b,f).32 on This TW2 is ,in Figurecorrelation 12c and with up tothe day measured 26 on TW3, RH readings, Figure 12 whiche). The show dry-out that ratethe RH during levels these reach periods saturation was with about the 1 g/(m2 h) (Figure 12d,f). This could be considered the maximum dry-out rate under these climate vapour· ‐retarding tape, making it comparable to a fully vapour sealed enclosure. With the vapour‐ conditions.permeable In tape, a later the phase RH oflevels the experiment,stabilised on when a higher the ambient level compared temperature to the was situation increased without to 31 ◦ Cany andadded then furthertapes. to 35 ◦C, the dry-out rate per square metre of the drying surface increased, which could be explained by the greater evaporation energy due to higher temperatures. However, the dry-out rate3.2. remained Simulation modest Results at best being 1.5–1.75 g/(m2 h). It is to be expected that with lower ambient · temperatures, the dry-out rate is lower. Unfortunately, due to overlapping experiments, it was not With laboratory measurements, it became evident that the moisture dry‐out capacity of the steel‐ possible to measure the dry-out rate when the TWs were placed in outdoor conditions. Nevertheless, faced sandwich panels is limited. The simulations were necessary to study the humidity regime of an the measured RH levels indicate that due to the high RH in the outside environment, the insulation entire sandwich panel over the course of three years with different initial moisture contents, joint core in the TWs quickly saturated with water vapour and in such conditions the dry-out capacity tapes, and solar radiation conditions. was negligible. The simulation results with a water vapour‐permeable tape (Sd 0.05 m) over the panel joint show The dry-out capacity was close to 0 g/(m2 h) with the vapour-retarding (S 15 m) tape (Figure 12d) that with the initial moisture content corresponding· to overhygroscopic level,d the outer layer of the and around 0.5 g/(m2 h) with the vapour-permeable (S 0.05 m) tape (Figure 12b,f). This is in stone wool insulation· core will experience water vapourd saturation conditions for more than 510 days correlation with the measured RH readings, which show that the RH levels reach saturation with if no solar radiation reached the surface of the sandwich panel (Figure 14a). It would take about two the vapour-retarding tape, making it comparable to a fully vapour sealed enclosure. With the years until the RH levels equalised at the levels produced if the initial RH had been 80% (Figure 14a). vapour-permeable tape, the RH levels stabilised on a higher level compared to the situation without The environment in the sandwich panel would become less humid with solar radiation on the any added tapes. surface of the wall (Figure 14b–d). Nevertheless, the RH in the centre of the panel would stay at water 3.2.vapour Simulation saturation Results level for more than 200 days if the wall were facing west (Figure 14c). This is 75 days more than if the initial RH were 80%. For the south direction, the initial humidity levels are comparableWith laboratory to those measurements, of the westward it becamefacing wall, evident but during that the the moisture following dry-out summers, capacity the RH of the level steel-faceddecreased sandwich more in panelsthe south is limited.‐facing Thewall simulations (Figure 14d). were However, necessary the to wall study with the the humidity initial regimemoisture ofcontent an entire on sandwich water vapour panel oversaturation the course level of still three needed years over with 10 di ﬀmonthserent initial to equalise moisture with contents, the RH joint levels tapes,of the and wall solar with radiation the initial conditions. RH of 80%. TheIt simulationis noteworthy results that with walls a water with vapour-permeablethe initial RH of 80% tape are (Sd also0.05 m)reaching over the the panel overhygroscopic joint show thatwater with limit the initialduring moisture wintertime content if the corresponding walls are not facing to overhygroscopic south, i.e., are receiving level, the little outer solar layer radiation. of the stoneThis wool is probably insulation due core to the will vapour experience‐permeable water vapourwind barrier saturation tape allowing conditions water for more vapour than from 510 outside days to enter the internals of the sandwich panel.

Sustainability 2020, 12, 9020 12 of 18 if no solar radiation reached the surface of the sandwich panel (Figure 14a). It would take about two

Sustainabilityyears until 20 the20, 1 RH2, x FOR levels PEER equalised REVIEW at the levels produced if the initial RH had been 80% (Figure12 of14 18a).

100 100

90 90

80 80

70 70

60 60 50 No solar radiation 50 Facing north

40 40 Relative humidity, % Relativehumidity, Jan Jul Dec Jul Dec Jul Dec % Relativehumidity, Jan Jul Dec Jul Dec Jul Dec Months P1_RH_sat P2_RH_sat P3_RH_sat Months P1_RH_sat P2_RH_sat P3_RH_sat P1_RH_80 P2_RH_80 P3_RH_80 P1_RH_80 P2_RH_80 P3_RH_80 (a) (b) 100 100

90 90

80 80

70 70

60 60 50 Facing west 50 Facing south

40 40 Relative humidity, % Relativehumidity, Jan Jul Dec Jul Dec Jul Dec % Relativehumidity, Jan Jul Dec Jul Dec Jul Dec Months P1_RH_set P2_RH_sat P3_RH_sat Months P1_RH_sat P2_RH_sat P3_RH_sat P1_RH_80 P2_RH_80 P3_RH_80 P1_RH_80 P2_RH_80 P3_RH_80 (c) (d)

FigureFigure 1 14.4. TwentyTwenty-eight-day-eight-day moving moving average average of of simulated simulated RH RH levels levels in in the the outermost outermost 3 3-mm-thick-mm-thick layerlayer of of the the mineral mineral wool wool core core at at various various distances distances from from the the panel panel joint joint (P1 (P1 being being closest closest to to the the joint, joint, P2P2 == atat one one-quarter-quarter of the panel; P3P3= = atat halfhalf of of the the panel) panel) and and with with an an initial initial RH RH of 80%of 80% and and at the at levelthe levelof water of water vapour vapour saturation saturation (dashed (dashed lines) oflines) a sandwich of a sandwich panel wall panel fully wall shaded fully (shadeda) or facing (a) or north facing (b), northwest ( cb)), or west south (c) (ord). south (d).

TheThe 28 environment-day moving in average the sandwich of simulated panel RH would levels become (Figure less 15 humid) in the withmiddle solar of the radiation panel in on the the outermostsurface of the3 mm wall thick (Figure layer 14 b–d). of the Nevertheless, mineral wool the core RH with in the varying centre of initial the panel moisture would contents stay at water and withoutvapour saturationany solar radiation level for more show than that 200 when days using if the the wall vapour were facing-permeable west (Figure(Sd = 0.05 14c). m) This tape is (Figure 75 days 1more5a), the than RH if the will initial stabilise RH were for 80%. most For initial the south moisture direction, content the levelsinitial humidity after onelevels year, are reaching comparable the overhygroscopicto those of the westward RH level facing in winter wall, but even during if the the initial following moisture summers, content the RH was level close decreased to 0 kg/m more3. Evidently,in the south-facing when using wall the (Figure vapour 14-retardingd). However, (Sd = the 15 wallm) tape, with the the outer initial layer moisture of the content insulati onon watercore willvapour remain saturation at approximately level still needed the level over of 10 overhygroscopic months to equalise moisture with the indefinitely RH levels ofif the the initial wall with RH theis moreinitial than RH 80% of 80%. (Figure 15b). The RH inside the sandwich panel did not reach the overhygroscopic level withIt is noteworthy lower initial that moisture walls with contents the initial correspon RH ofding 80% to are < also50% reachingRH during the the overhygroscopic simulated period, water butlimit the during model wintertime with an initial if the moisture walls are content not facing of 0 south, kg/m3 i.e.,showed are receiving a tendency little of solar rising radiation. RH. This is probably due to the vapour-permeable wind barrier tape allowing water vapour from outside to enter Joint tape S = 0.05 m Joint tape S = 15 m the100 internals of the sandwichd panel. 100 d 90The 28-day moving average of simulated RH levels90 (Figure 15) in the middle of the panel in the 80 80 outermost70 3 mm thick layer of the mineral wool core with70 varying initial moisture contents and without any60 solar radiation show that when using the vapour-permeable60 (Sd = 0.05 m) tape (Figure 15a), the RH 50 50 will40 stabilise for most initial moisture content levels after40 one year, reaching the overhygroscopic RH 3

level30 in winter even if the initial moisture content was30 close to 0 kg/m . Evidently, when using the Relative humidity, % Relative humidity,

20 % Relative humidity, 20 vapour-retardingJan0 183Jul (S365Detd = 15548Jul m) tape,730Det the outer913Jul layer1Det 095 of the insulationJan0 183Jul core365Det will remain548Jul 730Det at approximately913Jul 1Det 095 the levelMonths of overhygroscopicP3_RH_sat P3_RH_90 moisture indeﬁnitelyP3_RH_80 if theMonths initial RHP3_RH_sat is more thanP3_RH_90 80% (FigureP3_RH_80 15b). The RH inside the sandwich panel did not reach the overhygroscopic level with lower initial moisture P3_RH_50 P3_RH_0 P3_RH_50 P3_RH_0 contents corresponding(a) to < 50% RH during the simulated period, but(b the) model with an initial moisture content of 0 kg/m3 showed a tendency of rising RH. Figure 15. Twenty-eight day moving average of simulated RH levels in the outermost 3-mm-thick layer of the mineral wool core in the middle of the panel (output P3) with varying initial RH with a vapour-permeable (a) and a vapour-retarding (b) tape over the panel joint. Sustainability 2020, 12, x FOR PEER REVIEW 12 of 18

100 100

90 90

80 80

70 70

60 60 50 No solar radiation 50 Facing north

90 90

80 80

70 70

60 60 50 Facing west 50 Facing south

Figure 14. Twenty-eight-day moving average of simulated RH levels in the outermost 3-mm-thick layer of the mineral wool core at various distances from the panel joint (P1 being closest to the joint, P2 = at one-quarter of the panel; P3 = at half of the panel) and with an initial RH of 80% and at the level of water vapour saturation (dashed lines) of a sandwich panel wall fully shaded (a) or facing north (b), west (c) or south (d).

The 28-day moving average of simulated RH levels (Figure 15) in the middle of the panel in the outermost 3 mm thick layer of the mineral wool core with varying initial moisture contents and without any solar radiation show that when using the vapour-permeable (Sd = 0.05 m) tape (Figure 15a), the RH will stabilise for most initial moisture content levels after one year, reaching the overhygroscopic RH level in winter even if the initial moisture content was close to 0 kg/m3. Evidently, when using the vapour-retarding (Sd = 15 m) tape, the outer layer of the insulation core will remain at approximately the level of overhygroscopic moisture indefinitely if the initial RH is more than 80% (Figure 15b). The RH inside the sandwich panel did not reach the overhygroscopic levelSustainability with lower2020, 12 initial, 9020 moisture contents corresponding to < 50% RH during the simulated period,13 of 18 but the model with an initial moisture content of 0 kg/m3 showed a tendency of rising RH.

Joint tape S = 0.05 m Joint tape S = 15 m 100 d 100 d 90 90 80 80 70 70 60 60 50 50 40 40

30 30 Relative humidity, % Relative humidity,

20 % Relative humidity, 20 Jan0 183Jul 365Det 548Jul 730Det 913Jul 1Det 095 Jan0 183Jul 365Det 548Jul 730Det 913Jul 1Det 095

Months P3_RH_sat P3_RH_90 P3_RH_80 Months P3_RH_sat P3_RH_90 P3_RH_80 P3_RH_50 P3_RH_0 P3_RH_50 P3_RH_0 (a) (b)

FigureFigure 1 15.5. TwentyTwenty-eight-eight day day moving moving average average of of simulated simulated RH RH levels levels in in the the outermost outermost 3 3-mm-thick-mm-thick layerlayer of of the the mineral mineral wool wool core core in in the the middle middle of of the the panel panel (output (output P3) P3) with with varying varying initial initial RH RH with with a a Sustainabilityvapourvapour-permeable 20-permeable20, 12, x FOR ( (a aPEER)) and and REVIEWa a vapour vapour-retarding -retarding ( (b)) tape tape over over the the panel panel joint. joint. 13 of 18

TheThe calculated calculated time time of of wetness wetness (TOW) (TOW) in in the the middle middle of of the the panel panel in in the the out outermostermost 3 3-mm-thick-mm-thick layerlayer of of the the mineral mineral wool wool core core reaches reaches a a constant constant level level after after three three years years from from the the simulation simulation start start at at aboutabout 2000 h h/year/year when when using using the the vapour-permeable vapour-permeable (Sd =(S0.05d = m) 0.05 tape m) regardless tape regardless of the initial of the moisture initial moisturecontent (Figure content 16 (Figurea). TOW 16 woulda). TOW reach would to about reach 5400 to h about with initial 5400 h moisture with initial content moisture corresponding content correspondingto 80% RH or to more, 80% hadRH aor vapour-retarding more, had a vapour (Sd-retarding= 15 m) tape (Sd = been 15 m) used tape (Figure been used 16b). (Figure TOW 1 stays6b). below 200 h/year with vapour-retarding tapes and low initial moisture content ( 50%) but tends to rise TOW stays below 200 h/year with vapour-retarding tapes and low initial moisture≤ content (≤50%) buteach tends year. to rise each year.

Joint tape Sd = 0.05 m Joint tape Sd = 15 m 6000 6000

5000 5000 P3_RH_sat, 4000 4000 P3_RH_90 P3_RH_80 3000 3000

2000 2000 P3_RH_50,

P3_RH_0 TOW, h TOW, TOW, h TOW, 1000 1000

0 0 1 2 3 4 5 1 2 3 4 5 Time, years P3_RH_sat P3_RH_90 P3_RH_80 Time, years P3_RH_sat P3_RH_90 P3_RH_80 P3_RH_50 P3_RH_0 P3_RH_50 P3_RH_0 (a) (b)

FigureFigure 1 16.6. CalculatedCalculated TOW fromfrom thethe simulated simulated temperature temperature and and RH RH levels levels in thein the outermost outermost 3-mm-thick 3-mm- thicklayer layer of the of mineral the mineral wool wool core incore the in middle the middle of the of panel the panel (output (output P3) with P3) varying with varying initial initial RH with RH a withvapour-permeable a vapour-permeable (a) and (a a) vapour-retardingand a vapour-retarding (b) tape (b over) tape the over panel the joint.panel joint.

TheThe spatial distribution distribution of of RH RH in thein the sandwich sandwich panel panel on day on 183 day (summer) 183 (summer) illustrates illustrates how moisture how moisturewould be would distributed be distributed in the panel in the using panel a vapour-retarding using a vapour-retarding (Figure 17 (Figurea) or a 1vapour-permeable7a) or a vapour- permeabletape (Figure tape 17 (Figureb) onthe 17b) vertical on the vertical joint. Relativejoint. Relative humidity humidity becomes becomes evenly evenly distributed distributed with with the thevapour-retarding vapour-retarding tape. tape. This This suggests suggests that that the dry-outthe dry- isout negligible is negligible when when using using these these tapes. tapes. If the If RH the in RHthe panelin the had panel equalised had equalised to 80% before to 80% it was before sealed it with was asealed vapour-retarding with a vapour tape,-retarding the outermost tape, layer the outermostwould experience layer would conditions experience near theconditions overhygroscopic near the limitoverhygroscopic even during alimit warm even season during (Figure a warm 17a). seasonIf vapour-permeable (Figure 17a). If vapour tapes are-permeable used, the tapes RH levels are used, in the the panel RH levels will be in belowthe panel 90% will (Figure be below 17b) 90% and (Figuremostly 1 inﬂuenced7b) and mo bystly the inf outsideluenced climate. by the outside climate. However,However, this this means means that that with with the the vapour vapour-permeable-permeable tape, tape, the the panel panel also also acquires acquires excessive excessive moisturemoisture from thethe outsideoutside environment.environment. This This will will lead lead to to RH RH levels levels reaching reaching the overhygroscopicthe overhygroscopic limit limitduring during wintertime wintertime in the in outermost the outermost layers layers even if even the initialif the RHinitial were RH 50% were (Figure 50% 18(Figureb). This 18 doesb). This not doeshappen not whenhappen using when vapour-retarding using vapour-retarding tapes (Figure tapes 18 (Figurea). 18a).

Initial RH = 80%, joint tape Sd = 15 m

(a)

Initial RH = 80%, joint tape Sd = 0.05 m

(b)

Figure 17. Simulated spatial distribution of RH in the sandwich panel on day 183 (summer) with vapour-retarding (a) and vapour-permeable (b) tapes over the panel joint (the upper left corner in the graphs) with an initial RH of 80%. Sustainability 2020, 12, x FOR PEER REVIEW 13 of 18

The calculated time of wetness (TOW) in the middle of the panel in the outermost 3-mm-thick layer of the mineral wool core reaches a constant level after three years from the simulation start at about 2000 h/year when using the vapour-permeable (Sd = 0.05 m) tape regardless of the initial moisture content (Figure 16a). TOW would reach to about 5400 h with initial moisture content corresponding to 80% RH or more, had a vapour-retarding (Sd = 15 m) tape been used (Figure 16b). TOW stays below 200 h/year with vapour-retarding tapes and low initial moisture content (≤50%) but tends to rise each year.

Joint tape Sd = 0.05 m Joint tape Sd = 15 m 6000 6000

5000 5000 P3_RH_sat, 4000 4000 P3_RH_90 P3_RH_80 3000 3000

2000 2000 P3_RH_50,

P3_RH_0 TOW, h TOW, TOW, h TOW, 1000 1000

0 0 1 2 3 4 5 1 2 3 4 5 Time, years P3_RH_sat P3_RH_90 P3_RH_80 Time, years P3_RH_sat P3_RH_90 P3_RH_80 P3_RH_50 P3_RH_0 P3_RH_50 P3_RH_0 (a) (b)

Figure 16. Calculated TOW from the simulated temperature and RH levels in the outermost 3-mm- thick layer of the mineral wool core in the middle of the panel (output P3) with varying initial RH with a vapour-permeable (a) and a vapour-retarding (b) tape over the panel joint.

The spatial distribution of RH in the sandwich panel on day 183 (summer) illustrates how moisture would be distributed in the panel using a vapour-retarding (Figure 17a) or a vapour- permeable tape (Figure 17b) on the vertical joint. Relative humidity becomes evenly distributed with the vapour-retarding tape. This suggests that the dry-out is negligible when using these tapes. If the RH in the panel had equalised to 80% before it was sealed with a vapour-retarding tape, the outermost layer would experience conditions near the overhygroscopic limit even during a warm season (Figure 17a). If vapour-permeable tapes are used, the RH levels in the panel will be below 90% (Figure 17b) and mostly influenced by the outside climate. However, this means that with the vapour-permeable tape, the panel also acquires excessive moisture from the outside environment. This will lead to RH levels reaching the overhygroscopic Sustainabilitylimit during2020 wintertime, 12, 9020 in the outermost layers even if the initial RH were 50% (Figure 18b14). This of 18 does not happen when using vapour-retarding tapes (Figure 18a).

Initial RH = 80%, joint tape Sd = 15 m

(a)

Initial RH = 80%, joint tape Sd = 0.05 m

(b)

FigureFigure 17.17. SimulatedSimulated spatial spatial distribution distribution of ofRH RH in the in thesandwich sandwich panel panel on day on day183 (summer) 183 (summer) with withvapour vapour-retarding-retarding (a) and (a) andvapour vapour-permeable-permeable (b) tapes (b) tapes over over the thepanel panel joint joint (the (the upper upper left leftcorner corner in the in thegraphs) graphs) with with an initial an initial RH RHof 80%. of 80%. Sustainability 2020, 12, x FOR PEER REVIEW 14 of 18

Initial RH = 50%, joint tape Sd = 15 m

(a)

Initial RH = 50%, joint tape Sd = 0.05 m

(b)

FigureFigure 18. 18. SimulatedSimulated spatial spatial distribution distribution of of RH RH in thein the sandwich sandwich panel panel on day on day365 (winter) 365 (winter) with withvapour vapour-permeable-permeable (a) and (a )vapour and vapour-retarding-retarding (b) tapes (b) over tapes the over panel the joint panel with joint an with initial an RH initial of 50%. RH of 50%. 3.3. Critical Rain Amount and Moisture Safety Measures 3.3. Critical Rain Amount and Moisture Safety Measures The results from the laboratory test showed a limited dry-out capacity and the simulations indicatedThe results a high TOW from thelevel laboratory with vapour test-retarding showed tapes a limited and initial dry-out moisture capacity contents and the corresponding simulations indicatedto 80% RH a highor more. TOW The level difference with vapour-retarding in the moisture tapes content and initial correspo moisturending contents to 80% correspondingRH and water tovapour 80% RH saturation or more. level The diis ﬀveryerence small: in the 0.157 moisture kg/m3 contentor 3.61 × corresponding 10−2 kg for a panel to 80% with RH the and volume water vapour of 0.23 saturation level is very small: 0.157 kg/m3 or 3.61 10 2 kg for a panel with the volume of 0.23 m3. m3. Assuming the density of water to be 1000 kg/m×3 makes− the volume of this amount to be 3.61 × 10−5 Assuming the density of water to be 1000 kg/m3 makes the volume of this amount to be 3.61 10 5 m3. m3. This corresponds to roughly 1.57 × 10−4 m or ≈ 0.16 mm of water column over the surface× with− the This corresponds to roughly 1.57 10 4 m or 0.16 mm of water column over the surface with the dimensions of 0.23 m by 1 m. This× is a −remarkably≈ small amount. For triple the volume (e.g., had the dimensionswall enclosure of 0.23 been m 3 by m 1 high), m. This this is would a remarkably be ≈ 0.5 small mm. amount. Which means For triple that the a very volume small (e.g., amount had the of wall enclosure been 3 m high), this would be 0.5 mm. Which means that a very small amount of rain rain would be enough to raise the moisture≈ content to water vapour saturation levels inside the wouldsandwich be enough wall, had to the raise rain the ingressed moisture the content enclosure. to water For vapour example, saturation the mean levels precipitation inside the amount sandwich per wall,one wet had theday rain is 5.4 ingressed mm in the Estonia enclosure. [27], For and example, a rainfall the event mean is precipitation considered amount as heavy per when one wet the dayaccumulated is 5.4 mm sum in Estonia of precipitation [27], and a rainfallis over event50 mm is per considered day [28]. as In heavy a study when discussing the accumulated heavy rainfalls sum of in Estonia, the authors reported that in total 509 heavy precipitation events occurred in Estonia during a 44-year period [28]. Due to the combination of a limited dry-out and the possibility of a long time of wetness had a small amount of rain entered the panel, guidelines to enhance the moisture safety of such structures are presented in accordance with and in addition to the Swedish industry standard for moisture safety of the construction process [29].  A moisture management plan should be prepared considering the limited dry-out rate up to 2 g/day (without tapes, in favourable conditions) through a 30-mm-wide and 3-m-long joint.  The moisture management plan should also consider that the panels must not equalise to ambient air conditions if ambient RH > 80%. A protective film on the exposed sides of the insulating core must be in place from the factory if such conditions or rain might occur. The film should be removed before the installation of the panels.  If ambient RH > 80% during storage and installation, then vapour-permeable tapes must be used to cover the outer joints.  A moisture safety officer should be on site during the installation of the panels.  Moisture inspection rounds must be carried out regularly.  Spot measurements of humidity in the panels are recommended.  In the case of water ingress, all free water must be immediately removed from the structure and t&RH sensors must be installed where the leakage occurred. o If RH > 80% inside the panel, then aided drying is necessary and outer joints must not be covered with vapour-retarding tapes.  Non-conformances in relation to the moisture safety plan must be documented. Sustainability 2020, 12, 9020 15 of 18 precipitation is over 50 mm per day [28]. In a study discussing heavy rainfalls in Estonia, the authors reported that in total 509 heavy precipitation events occurred in Estonia during a 44-year period [28]. Due to the combination of a limited dry-out and the possibility of a long time of wetness had a small amount of rain entered the panel, guidelines to enhance the moisture safety of such structures are presented in accordance with and in addition to the Swedish industry standard for moisture safety of the construction process [29].

A moisture management plan should be prepared considering the limited dry-out rate up to • 2 g/day (without tapes, in favourable conditions) through a 30-mm-wide and 3-m-long joint. The moisture management plan should also consider that the panels must not equalise to ambient • air conditions if ambient RH > 80%. A protective film on the exposed sides of the insulating core must be in place from the factory if such conditions or rain might occur. The film should be removed before the installation of the panels. If ambient RH > 80% during storage and installation, then vapour-permeable tapes must be used • to cover the outer joints. A moisture safety officer should be on site during the installation of the panels. • Moisture inspection rounds must be carried out regularly. • Spot measurements of humidity in the panels are recommended. • In the case of water ingress, all free water must be immediately removed from the structure and • t&RH sensors must be installed where the leakage occurred.

If RH > 80% inside the panel, then aided drying is necessary and outer joints must not be # covered with vapour-retarding tapes. Non-conformances in relation to the moisture safety plan must be documented. • 4. Discussion The findings of the laboratory test and the following simulations indicate that the dry-out capacity of steel-faced sandwich panels is low (about 1 g/(m2 h)) even if there is no wind barrier tape covering · the vertical joint, which is the only area where moisture dry-out could occur at a significant level. A dry-out rate of 1 g/(m2 h) means that the moisture dry-out would be about 2 g/day through a · standard 30 mm wide and 3 m long vertical joint. The laboratory test demonstrated that when the vertical joint is covered with a vapour-permeable tape, the moisture dry-out will be reduced to a half of that of an uncovered joint (to 1 g/day). Moisture dry-out is also temperature driven and the dry-out rates presented here would appear under the most favourable conditions (high ambient temperature and low external RH). It is to be expected that with lower ambient temperatures, the dry-out rate is even lower. If the joint were covered with a vapour-retarding tape, no moisture dry-out is to be expected and in case of any water ingress into the insulation space, water vapour saturation would persist in the outer layers of the insulation core. This could have a negative impact on the glue adhering the steel sheet to the insulation core or to the steel sheet itself [2,13]. Pfeiffer reported the results of different temperature and humidity conditions on the tensile strength of a sandwich panel with mineral wool core. His research suggests that the tensile strength decreases the most when the sandwich panel experiences water vapour saturation and overall it is expected that increased humidity and temperature lead to a faster loss in strength [13]. Moreover, research shows that persisting elevated moisture content of mineral wool reduces its compressive strength and thermal insulation function significantly [14,17]. All of which leads to durability issues and lesser sustainability of the overall structures. In the simulations, a comparison was made between the situation with an initial moisture content corresponding to a water vapour saturated level and 80% RH. The results indicate that the vapour-permeable tape indeed allowed moisture to dry out; however, the process would be slow and it would take 10 to 24 months for the humidity levels to equalise with the levels produced by an initial Sustainability 2020, 12, 9020 16 of 18

RH level of 80% depending on the presence of solar radiation (from exposed directly towards south to a fully shaded situation). This suggests that moisture ingress could have a negative impact on the longevity of the structure regardless of the use of vapour-permeable tapes over joints. Nevertheless, the vapour-permeable tapes indeed help to clear out excess moisture from the panel. However, the simulations also indicate that the vapour-permeable tape allows vapour to enter the structure from outside and, regardless of the initial moisture content, an equilibrium state will arrive where there is an environment with high humidity levels over three months during the wintertime. This is something to be considered when using the vapour-permeable tapes of the joints—whether the yearly recurring high moisture levels are problematic or not. Further simulations demonstrated that when using the vapour-retarding tape, there was little vapour flow into the sandwich panel and with initial moisture contents corresponding to >50% RH, the humidity levels in the panel did not reach the overhygroscopic limit during the three-year simulation period. However, it was evident that even with the relatively vapour-retarding tape, the RH levels inside the panel still tended to rise. This suggests that over the course of the building lifespan, unfavourable conditions might occur. It is also common that these panels are installed during periods when the outside air RH is over 80% and if the panel had equalised to these conditions the outer layer of the insulation core would experience persisting water vapour saturation, leading to possible damage. TOW calculations indicated that regardless of the initial moisture content, TOW will stabilise to level T3 if vapour-permeable tapes are used. This suggests that if the airborne salinity contamination and the concentration of sulphur compounds in the sandwich panel are low, the corrosivity category of the inside environment of the sandwich panel would be up to C3 or medium [25]. In more aggressive environments or when using vapour-retarding tapes, the corrosivity category could reach C4. It is advisable to supply the panels sealed with a protective film over the exposed insulation core so that they would not adsorb moisture during the storage or installation process. This is to avoid the panel equalising to high RH (e.g., >80%) conditions, which could lead to persisting water vapour saturation in the external layer of the insulating core and a high TOW value. The protecting film should be removed right before the installation of the sandwich panel. The combination of limited moisture dry-out and the possibility of persisting water vapour saturation conditions in the outer layer of the insulation core due to small amount of rain leads us to suggest that the storage and installation process of such panels must be well managed. Other research also recommends the inclusion of moisture safety management in the building process to maintain a dry building process [11]. More research is needed to determine the limiting time of wetness or humidity criteria for such panels. A degradation model could be useful. The degradation model should consider corrosion, decomposition of the adhesive between steel and mineral wool, change of strength or thermal properties of the mineral wool and mould risk (if there is a possibility of air leakages to the indoor environment).

5. Conclusions This study investigated the moisture dry-out capacity and the hygrothermal regime of a steel-faced sandwich panel with mineral wool insulation. A series of laboratory tests and heat, air and moisture transport simulations were carried out. The results indicate that the moisture-dry out capacity of the studied sandwich panels is limited (up to 2 g/day through a standard 30-mm-wide and 3-m-long vertical joint without covering tapes in favourable conditions). When the vertical joint is covered with a vapour-permeable tape, the moisture dry-out will be reduced to a half of that of an uncovered joint (to 1 g/day) and if the joint were covered with a vapour-retarding tape, no moisture dry-out is to be expected. Use of vapour-retarding tapes on the outer joints will lead to persisting high humidity conditions and a high TOW value inside the panel if the initial relative humidity is higher than 80%. Vapour-permeable tapes allow moisture to dry out, but the rate of 1 g/day for a standard size panel leads to a time-consuming dry-out if there is any water leak into the panel. It is recommended that the sandwich panels be supplied on site in a sealed condition so that they would not adsorb moisture before installation and that, during the construction time, weather protection be used to avoid water Sustainability 2020, 12, 9020 17 of 18 ingress into the panels. The panels must not equalise to the outside air conditions if the air RH is 80% or higher. The inside to the steel sheets should withstand an environment with a corrosivity category at least C3 even with thorough moisture management and vapour-permeable tapes or C4 when using vapour-retarding tapes. The results indicate that very little moisture could lead to conditions where damages might occur. This means that the protection method should not allow water to enter the insulation space and in case of water ingress all free water must be immediately removed from the structure and additional aided drying is necessary if the RH in the panel exceeds 80%. This study conﬁrms that construction time moisture management is also important with structures that are regarded as moisture tolerant under usage conditions.

Author Contributions: Conceptualization, K.K. and T.K.; methodology, K.K. and T.K.; investigation, K.K. and H.P.; writing and visualization—original draft, K.K.; writing—review and editing, K.K. and T.K.; supervision, T.K.; funding acquisition, T.K. All authors have read and agreed to the published version of the manuscript. Funding: This research was supported by the EC—Research Programme of the Research Fund for Coal and Steel—HAIR—Improved Durability of Steel Sandwich Panel Constructions regarding Hygrothermal and AIRtightness Performance (grant No. 754185), Estonian Centre of Excellence in Zero Energy and Resource Efficient Smart Buildings and Districts, ZEBE (grant No. 2014–2020.4.01.15-0016) and funded by the European Regional Development Fund, by the Estonian Research Council (grant No. PRG483), and by the European Commission through the H2020 project Finest Twins (grant No. 856602). Conflicts of Interest: The authors declare no conflict of interest.

References

1. Khan, T.; Acar, V.; Aydin, M.R.; Hülagü, B.; Akbulut, H.; Seydibeyo˘glu,M.Ö. A review on recent advances in sandwich structures based on polyurethane foam cores. Polym. Compos. 2020, 41, 2355–2400. [CrossRef] 2. Davies, J.M. Lightweight Sandwich Construction; Davies, J.M., Ed.; Blackwell Science Ltd.: Oxford, UK, 2008; ISBN 9780470690253. 3. Hill, C.; Norton, A.; Dibdiakova, J. A comparison of the environmental impacts of different categories of insulation materials. Energy Build. 2018, 162, 12–20. [CrossRef] 4. Llantoy, N.; Chàfer, M.; Cabeza, L.F. A comparative life cycle assessment (LCA) of different insulation materials for buildings in the continental Mediterranean climate. Energy Build. 2020, 225, 110323. [CrossRef] 5. Peris Mora, E. Life cycle, sustainability and the transcendent quality of building materials. Build. Environ. 2007, 42, 1329–1334. [CrossRef] 6. Spitzner, M.H.; Koschade, R.; Cammerer, J. Entwicklung eines Sandwichelements mit Energie-Akkumulation, Energieverteilung und Dämmung (SEA). In FIW München Geschäfts-und Tätigkeitsbericht 2010; Forschungsinstitut für Wärmeschutz e.V. München: Gräfelfing, Germany, 2010. 7. Pihelo, P.; Lelumees, M.; Kalamees, T. Influence of Moisture Dry-out on Hygrothermal Performance of Prefabricated Modular Renovation Elements. Energy Procedia 2016, 96, 745–755. [CrossRef] 8. Kukk, V.; Külaots, A.; Kers, J.; Kalamees, T. Influence of interior layer properties to moisture dry-out of CLT walls. Can. J. Civ. Eng. 2019, 46, 1001–1009. [CrossRef] 9. Pihelo, P.; Kalamees, T. Commissioning of moisture safety of nZEB renovation with prefabricated timber frame insulation wall elements. Wood Mater. Sci. Eng. 2019, 1–8. [CrossRef] 10. EU 305/2011 Regulation of the European Parliament and the Council laying down harmonised conditions for the marketing of construction products and repealing Council Directive 89/106/EEC. Off. J. Eur. Union 2011, 54, 5–43. [CrossRef] 11. Mjörnell, K.; Arfvidsson, J.; Sikander, E. A Method for Including Moisture Safety in the Building Process. Indoor Built Environ. 2012, 21, 583–594. [CrossRef] 12. Olsson, P.; Tjäder, E. Suggestions for adjustment of ByggaF to improve the current use and suit the process of renovation. Energy Procedia 2017, 132, 921–926. [CrossRef] 13. Pfeiffer, L. Durability Assessment of Sandwich Panel Construction. Ph.D. Thesis, University of Surrey, Guildford, UK, 2005. 14. Ivaniˇc,A.; Kravanja, G.; Kidess, W.; Rudolf, R.; Lubej, S. The Influences of Moisture on the Mechanical, Morphological and Thermogravimetric Properties of Mineral Wool Made from Basalt Glass Fibers. Materials 2020, 13, 2392. [CrossRef][PubMed] Sustainability 2020, 12, 9020 18 of 18

15. EN 13523-26:2014—Coil Coated Metals—Test Methods—Part 26: Resistance to Condensation of Water; Technical Committee CEN/TC 139 “Paints and varnishes”; National Standards Authority of Ireland: Dublin, Ireland, 2014. 16. Karamanos, A.; Hadiarakou, S.; Papadopoulos, A.M. The impact of temperature and moisture on the thermal performance of stone wool. Energy Build. 2008, 40, 1402–1411. [CrossRef] 17. Jerman, M.; Cernˇ ý, R. Effect of moisture content on heat and moisture transport and storage properties of thermal insulation materials. Energy Build. 2012, 53, 39–46. [CrossRef] 18. Laukkarinen, A.; Vinha, J.; Kalbe, K.; Kesti, J.; Kalamees, T.; Honkakoski, E. Laboratory tests and modelling of mineral wool insulated steel sandwich panels. E3S Web Conf. 2020, 172, 17006. [CrossRef] 19. Greenspan, L. Humidity fixed points of binary saturated aqueous solutions. J. Res. Natl. Bur. Stand. 1977, 81, 89–96. [CrossRef] 20. Nicolai, A.; Grunewald, J.; Zhang, J.J. Recent improvements in HAM simulation tools: Delphin 5/CHAMPS-BES. In Proceedings of the 12th Symposium for Building Physics, Dresden, Germany, 29–31 March 2007; pp. 866–876. 21. Sontag, L.; Nicolai, A.; Vogelsang, S. Validierung der Solverimplementierung des Hygrothermischen Simulationsprogramms Delphin; Technische Universität Dresden: Dresden, Germany, 2013. 22. Kalbe, K.; Piikov, H.; Kesti, J.; Honkakoski, E.; Kurnitski, J.; Kalamees, T. Moisture dry-out from steel faced insulated sandwich panels. E3S Web Conf. 2020, 172, 17007. [CrossRef] 23. Kalamees, T.; Vinha, J. Estonian Climate Analysis for Selecting Moisture Reference Years for Hygrothermal Calculations. J. Build. Phys. 2004, 27, 199–220. [CrossRef] 24. EN 15026 Hygrothermal Performance of Building Components and Building Elements—Assessment of Moisture Transfer by Numerical Simulation; Technical Committee CEN/TC 89 “Thermal performance of buildings and building components”; European Committee for Standardization: Brussels, Belgium, 2007. 25. ISO 9223:2012—Corrosion of Metals and Alloys—Corrosivity of Atmospheres—Classification, Determination and Estimation. Available online: https://www.iso.org/standard/53499.html (accessed on 12 October 2020). 26. Hoseinpoor, M.; Prošek, T.; Babusiaux, L.; Mallégol, J. Toward more realistic time of wetness measurement by means of surface relative humidity. Corros. Sci. 2020, 177, 108999. [CrossRef] 27. Tammets, T.; Jaagus, J. Climatology of precipitation extremes in Estonia using the method of moving precipitation totals. Theor. Appl. Climatol. 2013, 111, 623–639. [CrossRef] 28. Mätlik, O.; Post, P. Synoptic weather types that have caused heavy precipitation in Estonia in the period 1961–2005. Est. J. Eng. 2008, 14, 195–208. [CrossRef] 29. ByggaF Industry Standard ByggaF—Method for Moisture Safety of the Construction Process. Branschstandard ByggaF—Metod för fuktsäker byggProcess, Version 08/05/2013 (English translation); 2013. Available online: http://www.fuktcentrum.lth.se/verktyg-och-hjaelpmedel/fuktsaekert-byggande/byggaf- metoden/ (accessed on 12 October 2020).

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional aﬃliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).