Improvement in Fire Resistance and Smoke Leakage Performance for Existing Polyvinyl Chloride Pipes Passing Through Walls

Abstract

Penetration points on walls must include firestop measures to prevent the spread of fire. Countries worldwide have established standardized testing protocols for firestop methods, and firestop products must pass relevant tests before they can be used in buildings. In the present study, a simple method was developed to enhance the smoke leakage performance and fire resistance of existing polyvinyl chloride (PVC) pipes passing through walls. Two sets of smoke leakage tests were performed, followed by two sets of fire tests. The smoke leakage tests were nondestructive and thus did not damage the specimens; consequently, the same specimens could be used in the smoke leakage and fire tests. The results of the smoke leakage tests indicate that the method using PVC pipes wrapped with galvanized steel sleeves outperforms the method without such wrapping. Moreover, the results of the fire tests suggest that galvanized steel sleeves considerably improve thermal insulation and safety. This finding is explained as follows: PVC pipes may burn and break apart at high temperatures, compromising fire compartmentation. A galvanized steel sleeve can retain a burning, broken pipe piece, thereby preventing injuries and stopping the detached pipe from continuing to burn on the ground.

1. Introduction

Pipelines and conduits in a building play roles comparable to those of the vascular and nervous systems in the human body. Conduits are distributed throughout a building, creating numerous penetration openings. Sites where conduits pass through walls or floors are generally referred to as “penetrations”, which often contain gaps that allow flames and smoke to travel from one side to another. Consequently, firestop measures must be implemented at penetrations to prevent the spread of fire [1]. Countries worldwide have established standardized testing protocols for firestop methods, and firestop products must pass relevant tests before they can be used in buildings. Examples of fire testing protocols include those described in the CNS 15814-1 [2], ISO 10295-1 [3], EN 1366-3 [4], and GB 23864 [5] standards, which are used to evaluate the thermal barrier rating (T-Rating) or flame barrier rating (F-Rating). Furthermore, the UL 1479 [6] and ASTM E814 [7] standards are used to evaluate the airtightness rating (L-Rating) in addition to the T-Rating and F-Rating. Polyvinyl chloride (PVC) pipes are frequently used as conduits in buildings because they are lightweight and corrosion-resistant, have lower fluid resistance compared with iron pipes, exhibit high mechanical strength, are easy to install and route, and entail relatively low construction costs. However, due to their ease of installation, PVC pipes are often routed through walls without pre-drilled openings or are installed without proper firestopping under the assumption that their small diameter minimizes the penetration area. This type of penetration is commonly found in older buildings in Taiwan, with conduits installed long before modern firestop regulations were established. To address this phenomenon, which raises concerns regarding compromised fire compartmentation integrity, the fire resistance and smoke leakage performance of the aforementioned type of penetration must be urgently enhanced. Smoke is a hazardous by-product during a fire, and it often causes casualties before flames reach the building occupants. Statistics on fire-related casualties in the United States in 2015 revealed that approximately 73% of fatalities resulted from smoke inhalation [8]. Similarly, fire reports in various European countries indicated that approximately 70% of fire casualties were attributable to smoke [9]. In Taiwan, a hospital fire that began in an operating room on the fourth floor produced rapidly spreading smoke, which rose to the floors above and caused injuries to 11 people [10]. Inadequate firestop measures can result in serious smoke-related threats. When PVC pipes are exposed to fire, they release various harmful gases, such as hydrogen chloride, dioxin-like compounds, carbon monoxide, toluene, and other pyrolytic by-products. Short-term inhalation of these substances may cause coughing, sore throat, dizziness, nausea, and breathing difficulties. Long-term exposure can lead to chronic respiratory conditions and an elevated risk of cancer.

In many existing buildings, PVC pipelines have already been installed. Under the condition that these PVC pipes are not to be removed or replaced, implementing a simple retrofit method to enhance the fire resistance and smoke-sealing performance at pipe penetrations would be an ideal solution. Standard firestopping methods typically involve certified penetration sealing systems to reinforce fire compartments; however, such approaches are often costly. For existing buildings, comprehensive upgrades may be limited due to budgetary constraints. Therefore, this study focuses on a simplified improvement technique for cases where PVC piping penetrates fire-rated compartments in existing structures. This method involves adding a metal sleeve over the PVC conduit to improve its fire resistance and smoke-sealing performance. Compared to conventional methods outlined in current firestop test specifications—such as fire collars, cuffs, and thermal insulation materials—this proposed approach is simpler and more feasible to implement. By applying a metal sleeve to the exterior of existing piping, this method enhances penetration performance without being dependent on the quality of firestop installation and also significantly reduces retrofit costs. Furthermore, this study presents a detailed method for on-site testing of smoke leakage performance at pipe penetrations. This study refers to penetration fire testing standards from various countries [1,2,3,4,5,6,7], the current conditions of existing buildings, and the fire resistance time requirements in the Building Technical Regulations [11], using full-scale testing to explore improvements in the fire performance of pipe penetrations. In addition, the smoke generated during fires is often the most fatal factor for occupants. As mentioned earlier, smoke spread is extremely dangerous; however, there is currently no method to determine the leakage volume of penetrations after construction is completed. In most countries, including Taiwan, only the fire resistance time for penetrations is regulated [11], while the smoke-sealing performance remains unaddressed. Focusing only on fire resistance while neglecting smoke leakage control at ambient temperatures is clearly inadequate. Moreover, this study refers to the on-site smoke control door testing method by Hung et al. [12] and proposes a new on-site testing application method for evaluating the smoke-sealing performance of penetrations. Currently, there is no relevant test method for the smoke-sealing performance of penetrations after installation, making it impossible to determine their actual performance. This new testing method, if widely adopted in the future, could become key to the effective implementation of smoke compartmentation, as the quality control of completed penetrations in actual construction is the practical focus. The principle of the smoke leakage performance test is based on a simplified one-dimensional flow assumption. It is assumed that a fire occurs on one side of the wall, generating pressure on the fire-exposed side. This pressure drives air through the gaps or openings in the wall, allowing it to reach the non-fire side. The airflow behavior can be derived from the law of conservation of energy using the Bernoulli equation. Therefore, by measuring the pressure difference between the two sides of the specimen and the air density during the test, the leakage rate of the specimen can be calculated. The testing method in this study follows international testing standards [1,2,3,4,5,6,7] and is not limited to use in buildings in Taiwan. The results of this study can also be applied to piping systems in buildings worldwide.

2. Experimental Details

2.1. Furnace

A high-temperature furnace was used in this study for fire generation. This furnace operates through the electronic ignition of liquefied gas, with diesel used as a supplementary fuel. With a depth of 1.2 m, the furnace can accommodate test specimens with a maximum height and width of 1.2 m each. The furnace walls are lined with ceramic fiber bricks with a thickness of 30 cm, and the furnace base is composed of thermal insulation bricks. The outer shell of the furnace consists of steel plates and a steel frame. The compensation wire used is of type WCA-H 4/0.65 × 2, with an outer layer wrapped in fiberglass. At the rear of the furnace, an exhaust vent is installed to discharge combustion gases, which are then routed through a chimney leading outdoors. For data acquisition, the system employs Japan Yokogawa data recorders. All instrument signals are initially sent to a DS 600 data recorder (Yokogawa, Yokogawa Taiwan Corporation, Japan), which processes and converts these signals before transmitting them to a DC 100 unit. The final electronic signals are then transferred to a computer through Cat. 6-grade Ethernet cables. The data recording interval is 10 s. Moreover, at half the height of the rear wall inside the furnace, a T-shaped pipe is installed, with its rear end connected to a differential pressure meter in the furnace. The pressure data are also transmitted to the DS 600 data recorder. Four thermocouples are installed inside the furnace, two of which function as controlled thermocouples. The furnace temperature is measured using K-type thermocouples, as illustrated in Figure 1. These thermocouples meet the CNS 5534 standard [13], ensuring a performance rating of at least 0.75. The thermocouple wires are encased in heat-resistant stainless steel tubing (16 gauge) with a diameter of 6.35 mm, which is further enclosed within an open-ended insulated stainless steel tube with an inner diameter of 14 mm. Each thermocouple junction protrudes 25 mm from the tube opening. To enhance temperature measurement sensitivity, all thermocouples inside the furnace are subjected to preconditioning at 1000 °C for 1 h prior to the first use [14]. The measurement accuracy of the thermocouples is ±3%.

2.2. Smoke Leakage Test Apparatus

The apparatus used in this study to perform smoke leakage tests was the same as that used by Hung et al. [12] to conduct on-site smoke leakage measurements for smoke doors. This apparatus is based on the measurement principle employed by Kuo et al. [15], was constructed in accordance with the CNS 15038 standard [16], and is capable of measuring the leakage rate of test specimens at ambient temperature. As illustrated in Figure 2, the apparatus consists of three major components: an air blower, a flow meter, and a test chamber. The air blower has a maximum airflow rate of 6.8 m³/min, a power rating of 1/4 HP, and an operating voltage of 110 V. It utilizes a three-phase electrical supply and has an outlet diameter of 50 mm. The blower speed is regulated by a variable-frequency drive, with a controllable frequency range of 0.01–650.00 Hz. The flow meter contains a Honeywell smart differential pressure transmitter (Honeywell International, Inc., Charlotte, NC, USA). This meter can be used to measure flow rates in the range of 0–75 m³/h with an accuracy of ±2.5%, and it supports fluid temperatures from −10 °C to 60 °C and a humidity level below 90%. The flow meter is installed between the outlet of the air blower and the test chamber, with both the inlet and outlet diameters of the flow meter being 50 mm. The test chamber comprises a walled structure and plastic sheeting, which is secured to the walls with airtight tape and strong adhesive. An opening is created on the sheeting to produce space for a Testo 510 pocket-sized differential pressure meter (Testo SE & Co. KGaA, Titisee-neustact, Germany). This meter can measure pressures in the range of 0–100 hPa with an accuracy of ±0.03 hPa. The apparatus is also equipped with a thermometer, hygrometer, and barometer. The thermometer measures temperatures ranging from −40 °C to +100 °C with a resolution of 0.1 °C, the hygrometer measures relative humidity values from 0% to 100% with a resolution of 0.1%, and the barometer has a measurement range of 300–1200 hPa and a resolution of 0.1 hPa.

2.3. Test Specimens

The specimens used in this study were walls constructed with 12 cm thick brickwork, and each wall had an opening with a diameter of 4 inches (10.16 cm). The pipes employed in this study were made of PVC, with a nominal diameter of 2 inches (5.08 mm), a wall thickness of 4.1 mm, and a length of 102 cm. A metal sleeve with a nominal diameter of 2.5 inches (6.35 cm) and a wall thickness of 3.0 mm was fabricated from galvanized steel. The PVC pipe and wall opening were packed with rock wool (density of 60 kg/m³). On the side that was not exposed to fire, the metal sleeve protruded 9.5 cm from the wall, whereas on the fire-exposed side, the PVC pipe protruded 30 cm from the wall. One end of the metal sleeve was welded to an iron plate with length, width, and thickness of 16, 8, and 0.5 cm, respectively. The iron plate had four pre-drilled holes for screw fixation to the wall (Figure 3). Two specimens were tested in this study. Because the smoke leakage test was nondestructive and caused no damage to the specimens, it was conducted before the fire test. First, a smoke leakage test was conducted on a PVC pipe without a sleeve (test A1-S), after which the pipe was subjected to a fire test (test A1-F). Subsequently, a smoke leakage test was conducted on a PVC pipe with a sleeve (test A2-S), after which the pipe was subjected to a fire test (test A2-F).

2.4. Test Procedure

2.4.1. Fire Tests

The 1 h fire tests in this study were conducted in accordance with the CNS 15814-1 standard [2], which is primarily based on the provisions of the ISO 10295-1 standard [3]. Both sets of guidelines define the same standard furnace temperature rise curve. The relationship between temperature and time for this curve is expressed as follows:

T = 20 + 345 log10(8t + 1)

where T is the average furnace temperature (°C), and t is the test duration (min).

The acceptable deviation (de) of the temperature–time curve is expressed as follows:

(1)

de ≦ 15%             5 < t ≦ 10;

(2)

de = 15 − 0.5 × (t − 10)%       10 < t ≦ 30;

(3)

de = 5 − 0.083 × (t − 30)%       30 < t ≦ 60;

(4)

de = 2.5%             t > 60;

where

de = (A − AS)/AS × 100

In the aforementioned equation, A is the area under the actual average furnace temperature–time curve (°C⋅min), and As is the area under the standard heating temperature–time curve (°C⋅min).

The furnace pressure was maintained at 20 ± 2 Pa in a horizontal plane located 100 mm below the PVC pipe. Temperature measurements were conducted 25 mm (T1) and 325 mm (T2) from the wall surface. According to CNS 15814-1 [2], a thermocouple must be installed at a position 25 mm from the wall. In addition, the standard requires that the extension length of the specimen on the unexposed side be at least 300 mm. Therefore, thermocouples are installed both before and after the wall penetration. If the temperature readings at both positions meet the fire resistance rating criteria, the specimen can be deemed to have satisfied the required fire resistance performance.

2.4.2. Smoke Leakage Tests

The smoke leakage tests were primarily based on the methodology proposed by Hung et al. [12]. Smoke leakage measurements at intermediate temperatures were excluded; however, measurements at ambient temperature were conducted in accordance with the CNS 15038 standard [16]. Each pipe specimen was tested thrice, and the average value was used for analysis. The test procedure is outlined as follows:

Step 1: Verification of the airtightness of the test chamber

First, both sides of the specimen were sealed with plastic sheeting by using airtight tape to ensure that the leakage rate of the test chamber was 0 under a pressure differential of 50 Pa. Subsequently, the airtightness of the plastic sheeting and specimen was verified using the soap bubble method, in accordance with the basic leakage rate requirements specified in the CNS 15038 standard [16].

Step 2: Measurement of the specimen leakage rate

The leakage rate of the specimen was measured under pressure differentials of 10, 25, and 50 Pa. The measured leakage rates were then adjusted to standard gas conditions (temperature of 20 °C (293.15 K) and standard atmospheric pressure (101,325 Pa)) in accordance with the formula in the CNS 15038 standard [16] to obtain the corrected leakage rate.

Each measured leakage rate of the specimen was corrected to standard gas conditions by using the following equation:

$${Q_a}^{\prime }={\displaystyle \frac{Q_a}{(T+273.15)}}\times \left[k\times (p_a+p_m)-3.795\times {10}^{-3}\times M_w\times p_{H_{2}O}\right]$$

(1)

where ${Q_a}^{\prime }$ is the leakage rate ($m^3/h$) of the specimen under standard gas conditions, $Q_a$ is the measured leakage rate ($m^3/h$) of the specimen at a temperature of $(T+273.15)$ and a pressure of $(p_a+p_m)$, $Q_b$ is the leakage rate ($m^3/h$) of the test chamber, $Q_t$ is the combined leakage rate ($m^3/h$) of the specimen and test chamber, and $k$ is a constant (293.15/101,325 = 2.89 × 10⁻³). Moreover, $T$ is the ambient temperature (°C), $p_a$ is the atmospheric pressure (Pa), $p_m$ is the pressure differential (Pa), $M_w$ is the relative humidity (%), and $p_{H_{2}O}$ is the saturation vapor pressure (Pa).

In the aforementioned step, the leakage rate of the test chamber ($Q_b$) was measured first. Subsequently, the specimen was installed in the chamber, and the combined leakage rate of the specimen and test chamber ($Q_t$) was measured. The parameter $Q_b$ was then subtracted from $Q_t$ to determine the actual leakage rate of the specimen ($Q_a=Q_t-Q_b$). Finally, the $Q_a$ value was converted into $Q_a\prime$ by using Equation (1).

Example Calculation

Assume that during testing, the atmospheric pressure is 102,000 Pa, the ambient temperature is 27 °C (300.15 K), the pressure increase is 10 Pa, the saturated water vapor pressure is 3567 Pa, and the relative humidity is 50%. The leakage rate under standard conditions can then be calculated as follows:

$$\begin{array}{l}{Q_a}^{\prime }={\displaystyle \frac{Q_a}{(T+273.15)}}\times \left[k\times (p_a+p_m)-3.795\times {10}^{-3}\times M_w\times p_{H_{2}O}\right]\\ ={\mathrm{Q}}_{\mathrm{a}}/(27+273.15)\times (2.89\times 10{10}^{-3})\times (102000+10)-3.795\times {10}^{-3}\times \left(3567\right)\\ =0.96 {\mathrm{Q}}_{\mathrm{a}}\end{array}$$

3. Results

Although only two specimens were tested in this study, the fire resistance test duration was 60 min. If the specimens had any issues, abnormal phenomena would have manifested during this period. In both tests, the furnace temperature followed the standard time–temperature curve specified in CNS 15814-1 [2], confirming the validity of the testing environment. Additionally, the specimens’ temperature rise during the 60-min test period was within reasonable expectations, indicating that the tests were successfully conducted. The results obtained are thus considered representative. Moreover, the method of installing metal sleeves is not a complex procedure: it simply involves attaching a sleeve over the PVC pipe, a task that can be performed by general plumbing technicians. Therefore, the test results from the two specimens are deemed reasonably representative, and extensive specimen testing is not deemed necessary.

3.1. Smoke Leakage Tests Result

Test A1-S was performed first. The PVC pipe and adjacent wall were sealed using airtight tape and plastic sheeting on both sides of the wall. After a differential pressure tube and an air blower were installed, the testing system was reset to 0. The pressure difference was then sequentially adjusted to 10, 25, and 50 Pa, with each differential maintained for 2 min. During this period, the measured volumetric leakage rate $Q_b$ was 0, indicating that the test chamber was installed appropriately and was airtight. The plastic sheeting behind the specimen was then removed to measure its volumetric leakage rate. On the day of testing, the ambient temperature was 26.4 °C, the relative humidity was 67%, the atmospheric pressure was 100,950 Pa, and the saturation vapor pressure was 3444.54 Pa. Equation (1) was used to convert the measured leakage rates to standard gas conditions, with the corrected leakage rates (${Q_a}^{\prime }$) for the specimen (referred to as specimen A1-S) being 211.2, 354.8, and 500.6 m³/h under pressure differentials of 10, 25, and 50 Pa, respectively. Test A2-S was conducted using the same procedure as that used for test A1-S. In test A2-S, the ${Q_a}^{\prime }$ values of the specimen (referred to as specimen A2-S) under standard gas conditions and pressure differentials of 10, 25, and 50 Pa were 9.85, 22.5, and 47.14 m³/h, respectively (Figure 4). These results indicate that with no leakage from the test chamber, the ${Q_a}^{\prime }$ values of specimen A2-S under pressure differentials of 10, 25, and 50 Pa were considerably lower than those of specimen A1-S. This result was obtained because the metal sleeve surrounding the PVC pipe in specimen A2-S allowed for the insertion of additional rock wool between the sleeve and the PVC pipe, making leakage less likely. By contrast, in the A1-S specimen, only the gap between the PVC pipe and the 12 cm thick brick wall was filled with rock wool. Completely packing this gap is difficult in practice, and incomplete packing results in the formation of pathways for gas leakage.

3.2. Fire Tests Result

The criteria for determining fire resistance were as follows:

• During the 1 h heating test, no openings formed that could lead to the ignition of the cotton pad or allow the penetration of the measurement gauge.
• During the 1 h heating test, no sustained flames lasting longer than 10 s appeared on the side unexposed to the fire.
• During the 1 h heating test, the average temperature on the unexposed side did not exceed the initial average temperature of 140 °C, and the temperature at any given point did not exceed the initial average temperature of 180 °C.

The fire resistance performance classification is as follows: an F-Rating of 60 min and a T-Rating of 60 min.

The first fire test (test A1-F) was conducted on the PVC pipe without a sleeve (1 h fire resistance assessment). The initial furnace temperature was 29 °C, and the permissible average temperature on the unexposed side was 169 °C, with the temperature at any single point not to exceed 209 °C. At 35 min, the temperature at measurement point T1 had already reached 210.4 °C. Moreover, the temperature at T2 reached 181.4 °C. Both temperatures exceeded the limits set by the CNS 15814-1 standard [2], and the specimen exhibited sustained flames for longer than 10 s (Figure 5). The PVC pipe on the fire-exposed side burned through and fell to the ground (Figure 6), and ignition occurred. Consequently, the fire resistance was determined to be 32 min. Considering that Taiwan’s regulations require a fire resistance of at least 1 h, test A1-F was deemed unsuccessful. During this test, the furnace temperature conformed to the standard temperature rise curve. The specimen temperature during the test is illustrated in Figure 7. The furnace was allowed to cool to ambient temperature the next day after the first fire test, and the test site was cleared. Subsequently, the second fire test (A2-F) was conducted on the PVC pipe with a metal sleeve (1 h fire resistance assessment). The initial furnace temperature was 30 °C, and the permissible average temperature on the unexposed side was 170 °C, with the temperature at any single point not to exceed 210 °C. The temperature at T1 increased with time, reaching a maximum of 173.1 °C by the end of the test. Moreover, the temperature at T2 reached a maximum value of 103.2 °C. The specimen did not exhibit sustained flames exceeding 10 s and did not exceed the thermal barrier temperature thresholds specified in the CNS 15814-1 standard [2]; therefore, the PVC pipe on the unexposed side did not fall to the ground (Figure 8). The furnace temperature remained consistent with the standard temperature rise curve in test A2-F. The specimen temperature during this test is depicted in Figure 9. The aforementioned results indicate that the metal sleeve around the PVC pipe used in test A2-F successfully improved the pipe’s fire resistance, with the pipe satisfying the requirement for 1 h fire resistance.

4. Discussion

In the smoke leakage tests, specimen A2-S exhibited lower corrected leakage rates (${Q_a}^{\prime }$) than specimen A1-S under all tested pressure differentials. This result was obtained primarily because the metal sleeve around the PVC pipe (specimen A2-S) allowed additional rock wool to be packed into the gap between the sleeve and the pipe, which impeded gas flow. According to Ye, Z. et al. [1], sufficiently thick rock wool packing can effectively reduce smoke leakage. The leakage rate of specimen A2-S was up to 90% lower than that of specimen A1-S, demonstrating that wrapping a PVC pipe in a metal sleeve enhances the pipe’s smoke leakage performance. Many designers only consider fire resistance requirements when designing penetration points. However, ambient-temperature smoke leakage performance is also a critical parameter [17]. During fires, smoke-related fatalities often exceed those caused directly by flames [18]. This phenomenon occurs because smoke can travel through ceiling gaps and reach a penetration long before flames do [1,19]. Although the temperature of smoke is only marginally higher than room temperature at this stage, its accumulation can still prove fatal. Furthermore, as smoke layers descend, they obstruct evacuation efforts [20]. Therefore, improving the smoke leakage performance of penetration points is crucial. In test A1-F, the temperature at measurement point T1 reached 210.4 °C in 35 min, thereby exceeding the threshold of the CNS 15814-1 standard [2]. By contrast, in test A2-F, the maximum temperature at T1 was only 173.1 °C after 60 min. In addition to meeting the requirement of 1 h fire resistance, the specimen used in A2-F exhibited considerably higher thermal insulation than that used in A1-F, which showed fire resistance for only approximately 35 min. The installation of a metal sleeve not only improves thermal insulation but also enhances safety. For example, PVC pipes can melt at high temperatures, compromising fire compartmentation. By contrast, a metal sleeve can keep a melted PVC pipe contained within it, reducing the risk of injury and preventing the burning pipe from falling to the ground. In the present study, only the unexposed side of the PVC pipe was fitted with a metal sleeve. Nevertheless, the sleeve still considerably improved the pipe’s fire resistance. However, because a fire can occur on either side of a wall, metal sleeves are recommended to be installed on both sides of the wall. In the present study, the metal sleeve was inserted directly from the pipe end; however, such installation may be feasible only for short PVC pipes in a test environment. In actual buildings, conduits can be quite long; therefore, cutting PVC pipes and inserting metal sleeves may not be practical. Consequently, splitting a metal sleeve in half for installation could be a more feasible approach, provided that the two parts are appropriately anchored to the wall. Given that the primary function of a metal sleeve is to support a PVC pipe and accommodate rock wool, a half-sleeve design may be adequate. Finally, a schematic view of the galvanized sleeve cut in half at the midpoint is shown in Figure 10. Although other tests were performed during fire resistance testing in this study—such as testing with metal sleeves installed on both the fire-exposed and non-fire-exposed sides of the specimen—it was eventually decided to present only the method where the metal sleeve was installed on the non-fire-exposed side. The reason is that if installing the metal sleeve only on the non-fire side can already pass the fire resistance test, then in actual buildings, if metal sleeves are installed on both sides of the wall, the result should at least reach the performance demonstrated in this study. This study focuses solely on PVC pipes, which are commonly used in Taiwan, and confirms the effectiveness of the proposed improvement. If other pipe materials or sizes are involved in the future, further relevant testing is recommended. The method proposed in this study should improve the fire resistance performance of PVC pipes, as PVC itself is not fire-resistant and tends to melt. Therefore, the addition of a metal sleeve can provide at least some level of enhancement. The thickness of the metal sleeve used in this study is 3.0 mm. Generally, galvanized steel of this thickness offers good durability and corrosion resistance and can last for a long time. However, if corrosion issues are a concern in specific environments, it is recommended to use stainless steel sleeves instead. The smoke leakage test used in this study can be applied in future assessments of the ambient-temperature smoke leakage performance of penetrations after installation. This could serve as a basis for the acceptance of penetrations, as smoke-blocking performance at ambient temperatures should not be overlooked. This study focuses on improving the fire resistance performance of existing PVC pipes. Firestopping methods available on the market can achieve the same effect since they are all tested in accordance with CNS 15814-1 [2]. However, the method proposed in this study allows for immediate functionality, is easier to install, and reduces improvement costs. In comparison, existing methods require thorough sealing to be effective, making the method presented in this study simpler.

5. Conclusions

The conclusions of this study are as follows:

• In smoke leakage tests, a PVC pipe with a metal sleeve (specimen A2-S) exhibited considerably lower leakage rates than did a PVC pipe without a metal sleeve (specimen A1-S) under pressure differentials of 10, 25, and 50 Pa. This result is attributable to the metal sleeve, which allowed for a larger quantity of rock wool to be packed into the gap between the sleeve and the PVC pipe, which effectively prevented gas flow to the other side. The leakage rate of specimen A2-S was up to 90% lower than that of specimen A1-S.
• In fire resistance tests, specimen A1-S had a fire resistance duration of approximately 35 min, whereas specimen A2-S had a fire resistance duration exceeding 60 min. Thus, the metal sleeve enhances the fire resistance of the PVC pipe. The metal sleeve also contributed to improved safety. PVC pipes exposed to high temperatures in a fire may melt, thus compromising fire compartmentation. The presence of a metal sleeve helps contain the melted part of the PVC pipe within the sleeve, preventing the pipe from falling and causing injuries or igniting other materials.
• In this study, the metal sleeve was installed only on the side of the wall that was not exposed to the fire; nevertheless, the sleeve still considerably improved the smoke leakage performance and fire resistance of the PVC pipe. However, because a fire can occur on either side of a wall, metal sleeves are recommended to be installed on both sides of the wall in practical scenarios.
• Metal sleeves are recommended to be installed using a half-sleeve design. The primary function of a metal sleeve is to support a PVC pipe and accommodate rock wool insulation. Therefore, as long as the two parts of the metal sleeve are securely fixed to the wall, a half-sleeve design is feasible in practical construction.
• The smoke leakage test used in this study can be used to evaluate and ensure the integrity of smoke compartmentation. Moreover, this test can be employed to assess the ambient-temperature smoke leakage performance of penetrations after installation.