Hybrid Dipping–ShowerWet Scrubber for Multi-Pollutant Emission Reduction From Marine Diesel Exhaust: Experimental Assessment of a Low-Cost Prototype as an Auxiliary Exhaust Gas Cleaning Approach

The International Maritime Organization (IMO) adopted its Initial Strategy on the reduction of greenhouse gas (GHG) emissions from ships in April 2018 (Resolution MEPC.304(72)), which aimed to reduce the carbon intensity of international shipping by at least 40% by 2030 and its total annual GHG emissions by at least 50% by 2050, both relative to 2008 [1]. In July 2023, the Initial Strategy was replaced by the 2023 IMO Strategy (Resolution MEPC.377(80)), which raised the ambition to net-zero GHG emissions by or around 2050 and introduced indicative checkpoints of at least a 20% reduction (striving for 30%) by 2030 and at least a 70% reduction (striving for 80%) by 2040, relative to 2008 [2], [3]. Modelling of the 2023 targets indicates that they are broadly consistent with a well-below 2 °C trajectory but fall short of what is required to limit warming to 1.5 °C [4]. Because the eventual contribution of shipping to global warming depends on the cumulative emissions released until the sector reaches net zero, the pace of near-term reductions matters as much as the 2050 end point [5]. Figure 1 presents cumulative well-to-wake GHG emissions of international shipping from 2020 to 2050 for the baseline, IMO Minimum, IMO Striving and 1.5 °C pathways. The bars separate unavoided emissions and avoided emissions from operational efficiency and GFI standards, while dashed lines denote the 2 °C, 1.7 °C, and 1.5 °C carbon budgets [4].

Strategies and preventive measures to reduce GHG emissions have been implemented by the IMO in short, medium, and longterm programs. These measures are being taken to achieve the 2030 and 2040 targets [3]. The exhaust of a ship’s main engine contains both greenhouse gases and air pollutants. Carbon dioxide (CO$_{2}$) is the principal greenhouse gas, whereas sulfur dioxide (SO$_{2}$) and nitrogen oxides (NO$_{x}$ = NO + NO$_{2}$) are air pollutants rather than direct greenhouse gases: SO$_{2}$ forms sulfate aerosols that cause acidification and exert a net cooling rather than warming effect, and NO$_{x}$ contributes to ozone and secondary particle formation and to acid deposition. NO$_{x}$ should not be confused with nitrous oxide (N$_{2}$O), which is a greenhouse gas [6].

Effective emission control technology is necessary to lessen the detrimental effects of the shipping industry’s substantial contribution to environmental pollution through exhaust emissions from the burning of fossil fuels in ship engines [7]. One of the Exhaust Gas Cleaning Systems technologies that is frequently used to lower sulfur dioxide emissions to the 0.5% threshold in compliance with international regulations is wet scrubbers [8], [9]. In order to capture SO$_{x}$ and solid particulates prior to their release into the atmosphere, the Wet Scrubber system directs exhaust gases through high-pressure seawater or freshwater spray [10]. The alkalinity of seawater or other chemicals in the freshwater system neutralize the sulfuric acid that is created chemically from a sulfur dioxide solution in the wet scrubber process [11], [12]. To guarantee adherence to emission regulations, this system typically consists of residue handling facilities, wash water treatment installations, scrubber units positioned around the stack, and continuous emission monitoring systems [13]. There are three primary categories of wet scrubber systems: open loop, which uses seawater as a scrubbing medium; closed loop, which uses fresh water mixed with sodium hydroxide; and hybrid systems, which combine the two [14], [15]. This hybrid system offers high operational flexibility as it allows switching between open loop and closed loop modes depending on water conditions or local emission regulations [16], [17].

Generally, standard wet scrubbers have a negligible effect on CO$_{2}$ removal from ship exhaust [18]. This is primarily due to the low solubility of CO$_{2}$ in the aqueous solutions typically used in these scrubbers, such as seawater [18]. Wet scrubbing processes utilizing certain urea solutions have shown less than 10% CO$_{2}$ absorption efficiency [19].

Wet scrubbers demonstrate high efficiency in removing SO$_{x}$ from ship exhaust. Removal rates often exceed 95%, with some studies reporting up to 98% reduction [20]. Freshwater scrubbers, for instance, are capable of reducing SO$_{x}$ by up to 97% [12], [15]. To achieve emission levels comparable to using 0.1% sulfur fuel with a 3.5% sulfur fuel, scrubbers need to remove approximately 97.1% of SO$_{x}$ [21]. Electrified scrubbers can further enhance the absorption efficiency of sulfur dioxide [18].

In general, nitrogen oxides, such as nitric oxide and nitrogen dioxide, cannot be effectively reduced by wet scrubbers [22]. The primary issue is that nitric oxide, which makes up a significant amount of NO$_{x}$ emissions (approximately 90%), is difficult to scrub and has very low water solubility [23]. Technologies like Selective Catalytic Reduction are commonly used to significantly reduce NO$_{x}$ in marine exhaust. To effectively address both SO$_{x}$ and NO$_{x}$ emissions, these systems are frequently coupled with scrubbers [24].

Wet scrubbers are also utilized to lower exhaust outlet operating temperatures. The purpose of wet scrubber systems is to drastically lower the temperature of gas emissions. For instance, in the primary quenching stage of a multi-stage scrubbing process, gas temperatures can decrease from roughly 900 °C to 450 °C [25]. This temperature drop is relevant for two distinct reasons. First, when the heat removed from the exhaust is recovered upstream of the scrubber, for example in an exhaust-gas economizer or heat-recovery unit, it can raise overall boiler or plant efficiency [26]; simply quenching the gas inside a scrubber, as in the present prototype, removes this heat without recovering it. Second, a lower gas temperature increases the solubility of SO$_{2}$ in the scrubbing liquid and therefore favours its absorption [27].

There is a large research gap in the thorough evaluation and optimization of sophisticated wet scrubber technologies, such as the shower and dip wet scrubber, for the efficient removal of various contaminants from marine diesel exhaust under dynamic marine operating conditions. This gap underlines the significance of resolving crucial elements such as hydrodynamics, mass transfer, solvent stability, long-term performance, and realistic industrial scale deployment.

Current literature underscores a substantial necessity for interdisciplinary research to assess the cleaning efficacy, economic implications, and environmental consequences of alternative emission reduction technologies, given the ambiguous environmental advantages of scrubbers [11]. There exists a significant performance disparity between theoretical modeling and actual in-situ implementation that necessitates resolution to validate results across various ship typologies [28]. Moreover, current research indicates that wet scrubbers possess limited efficacy in mitigating the health and environmental effects of ship emitted particulates, while the transference of pollutants like PAHs into marine ecosystems through effluent discharge demands a thorough assessment of both air and sea interfaces [17], [29]. Presently, there is a lack of comprehensive, measurement-informed life cycle assessments that evaluate the potential environmental impacts of heavy fuel oil and scrubber production and operation in comparison to low-sulfur fuels [30]. Another major technical problem is that artifact nanoparticles form in high humidity and sulfur environments. This needs new engineering solutions that work well for both SO$_{X}$ and particulate abatement [17]. As a result, future studies should focus on creating integrated assessment frameworks that measure atmospheric dispersion, marine water quality degradation, and human health outcomes at the same time in order to find out how cost effective shower and dip wet scrubber technologies really are [29].

Despite this body of work, single-mode wet scrubbers exhibit intrinsic limits for multi-pollutant treatment. Immersion (dipping or bubble-column) scrubbers provide long gas residence time but a comparatively small interfacial area that is fixed by bubble size, whereas spray (shower) scrubbers create a very large droplet surface area but suffer from short gas–liquid contact time and droplet entrainment at high gas flow rates. Each mechanism therefore favors a different subset of pollutants and operating conditions, and neither alone performs consistently across the variable engine speeds typical of marine diesel operation.

Moreover, although combined spray-and-bubble concepts have been demonstrated at laboratory and pilot scale for flue gas, experimental data on a structurally combined “dipping + shower” contact pattern applied to small marine diesel (or simulated marine diesel) exhaust under varying engine speeds are still lacking. It remains unclear how much additional removal a combined contact pattern actually delivers for each gas (SO$_{2}$, CO, CO$_{2}$, NO$_{2}$), and how stable that performance is across the engine operating range.

This study addresses that gap through four specific contributions: (i) the design and construction of a low-cost prototype hybrid wet scrubber, fabricated largely from recycled industrial oil-storage drums, that integrates a dipping (bubble-column immersion) subsystem and a shower (spray atomization) subsystem in one unit; (ii) a systematic comparison of three operating modes (dipping-only, shower-only, and hybrid); (iii) simultaneous multi-pollutant measurement of SO$_{2}$, CO, CO$_{2}$, and NO$_{2}$ at the scrubber inlet and outlet; and (iv) performance testing at four engine-speed settings (25%, 50%, 75%, and 100% of rated speed). It should be emphasized that the term “hybrid” in this paper refers to this structural combination of bubble-immersion and spray gas--liquid contact within a single scrubber; it does not refer to the open-loop/closed-loop switching systems that are conventionally called hybrid scrubbers in the marine sector.

The R&D (Research & Development) methodology is a methodical, iterative process used in education, technology, and industry to bridge theory and practice. It usually entails needs analysis, design, prototyping, testing (validation), revision, and small scale implementation [31].

Figure 2 is the schematic diagram of the hybrid wet scrubber prototype coupled to a Dongfeng diesel engine. Exhaust gas enters via a submerged dip pipe in the dipping zone, rises through the water column and demister, passes the overhead spray zone, and exits from the top outlet. Circled numbers represent valves 1–8; S1 and S2 correspond to the inlet and outlet sampling points, and the dashed blue line indicates the operating liquid level.

The wet scrubber is designed using many recycled materials from the industrial world, one of which is the use of used oil storage drums as the main body or tube of the wet scrubber system. This wet scrubber uses two gas washing systems: Scrubber system 1 uses the dipping wet scrubber method, where the exhaust silencer mounting is submerged in water, trapping the gas emissions in the water medium. Scrubber System 2 uses the shower wet scrubber method, where the gas coming out of the exhaust silencer is not immersed in water but is only sprayed with water coming out of the shower mounting that has been installed in the scrubber system, while Scrubber System 3 is a combination of scrubber systems 1 and scrubber systems 2.

Scrubber system 1 uses a depth in water system, which directs exhaust gases or pollutants into a volume of water. The exhaust gases are directed into the volume of water through a submerged Ø2 in inlet (dip) pipe that discharges near the vessel base and acts as a simple sparger; no Venturi throat is present. This process is also a method of trapping and absorbing exhaust pollutants through a static volume medium. The scrubber system 2 uses the Splashing Water system, which utilizes water splashes or sprays fed into the scrubber system filled with exhaust gas from Dongfeng diesel engines. These water splashes act as a medium for capturing and absorbing pollutant gases produced by diesel engines.

In all configurations, the exhaust gas follows the same path through the vessel (Figure 2): it leaves the silencer through a Ø2 in pipe that descends into the scrubber and discharges near the vessel base, rises through the water column (dipping zone), passes the demister pad, crosses the spray zone beneath the overhead shower header, and leaves through the Ø2 in outlet at the top. The three configurations were obtained with the valves shown in Figure 2: dipping-only, shower supply valve 1 closed and the vessel filled to the overflow level set by valve 7; shower-only, vessel drained through valve 6 below the outlet of the dip pipe so that the gas is not immersed, with valve 1 open; and hybrid, vessel filled to the overflow level with valve 1 open. Valve 2 supplies make-up water directly to the dipping zone when the liquid level needs topping up. The inlet sampling point (S1) was located at the silencer, upstream of the scrubber, and the outlet sampling point (S2) at the Ø2 in scrubber outlet pipe; both points, the operating liquid level, and the make-up water line are marked in Figure 2.

In this paper, the treatment given to exhaust gas emissions from diesel engines was measured directly using the IoT-based multi-gas sensor node without a scrubber at varying engine-speed settings. The detailed experimental setup and operating conditions are summarized in Table 1. Furthermore, in the next experiment, exhaust gas emissions from the same diesel engine were passed through dipping and shower wet scrubber method, as well as a combination of both. In the first scenario, the engine system is turned on without using a scrubber, so that the measured exhaust emissions are the original emission values from the diesel engine exhaust through four different engine-speed settings, namely 25%, 50%, 75%, and 100%. In the second scenario, the engine system is turned on using scrubber system 1 so that the measured exhaust emissions are the emission values resulting from the scrubber system that has been created and acquired through four different engine-speed settings, namely 25%, 50%, 75%, and 100% of the rated speed of 2,200 rpm (i.e., 550, 1,100, 1,650, and 2,200 rpm). In the third scenario, the engine system is turned on using scrubber system 2 so that the measured exhaust emissions are the emission values resulting from the processing of the scrubber system that has been created and acquired through four different engine-speed settings, namely 25%, 50%, 75%, and 100% of rated speed. In the fourth scenario, the engine system is turned on using a combination of scrubbers 1 and 2 so that the measured exhaust emissions are the emission values resulting from the combined scrubber system that has been created and acquired through four different engine-speed settings.

Gas concentrations at the scrubber inlet and outlet were measured with a purpose-built IoT-based multi-gas sensor node. The node consists of an ESP32 microcontroller with an I²C multiplexer reading five gas sensors: a non-dispersive infrared (NDIR) MH-Z16 sensor for CO$_{2}$ (range 0–50,000 ppm, UART output, integrated temperature compensation) and factory-calibrated DFRobot electrochemical sensors for SO$_{2}$ (0–20 ppm), NO$_{2}$ (0–20 ppm), CO (0–1,000 ppm), and O$_{2}$ (single-point calibrated in ambient air at 20.9% vol). The electrochemical sensors apply built-in temperature-compensation algorithms, and the whole node was warmed up for at least five minutes (three minutes for the NDIR sensor) before each test. Sensor readings were streamed over MQTT (Mosquitto broker on a Raspberry Pi 5) and logged and visualized in Node-RED. The exhaust gas entered the scrubber at approximately 90 °C and was cooled to 30–40 °C at the outlet; the exhaust volumetric flow rate was not measured, which is acknowledged as a limitation when interpreting flow-related trends. At each engine-speed setting, the engine was first allowed to reach steady operation at the set speed before data acquisition, after which five consecutive readings were taken for each gas at both sampling points from the data logged during the 60-min run. The time interval between these readings was not recorded in the dataset, which is acknowledged as a limitation. All efficiencies in this paper are reported as the mean of these five readings. Because consecutive readings from the same run may be autocorrelated, their standard deviation (SD) and coefficient of variation (CV) describe short-term measurement variability within a single run; they are not an estimate of the uncertainty that five independent experimental replicates would provide. It should be noted that this is a preliminary prototype test based on a single experimental campaign; independent replication of the whole test series was not performed, and the limited statistical support is acknowledged as a limitation of this study.

The electrochemical SO$_{2}$ and NO$_{2}$ sensors have a full-scale range of 20 ppm and the CO sensor of 1,000 ppm, and their output is capped at full scale. Of the 240 inlet readings, 34 were recorded at full scale (SO$_{2}$: 9, all at 100% of rated speed; NO$_{2}$: 22; CO: 3, all at 100% of rated speed), and 11 NO$_{2}$ outlet readings (all at 75% and 100% of rated speed) were also at full scale; no value above the range could be registered. These readings were retained as recorded, without correction or exclusion. Because a saturated inlet reading can only underestimate the true inlet concentration while the corresponding SO$_{2}$ and CO outlet readings were below full scale, the SO$_{2}$ and CO efficiencies at 100% of rated speed should be regarded as lower-bound estimates. For NO$_{2}$, both inlet and outlet readings were saturated in several cases, so the NO$_{2}$ efficiencies, particularly at 75% and 100% of rated speed, are not quantitatively reliable. Where a mean + SD error bar in Figure 3, Figure 4, and Figure 5 extends beyond the full-scale value, this reflects the statistics of the recorded values, not a measured concentration above the sensor range.

The analysis and evaluation process was conducted comprehensively through a direct measurement approach. The direct measurement approach was used to determine the extent to which scrubbers can reduce the percentage of exhaust pollutants from diesel engines. This performance parameter indicates the scrubber’s ability to absorb gas pollutants produced by diesel engines. The more gas pollutants absorbed by the scrubber, the higher the efficiency of pollutant gas absorption. The efficiency percentage shows the comparison between the average emission value with a scrubber and the average emission value without a scrubber.

In this study, the concentrations of CO$_{2}$, SO$_{2}$, NO$_{2}$, and CO were recorded at both the inlet and the outlet of the scrubber, following the inlet--outlet comparison approach used in previous scrubber studies [30-32]. Particulate matter, NO, and total SO$_{x}$ were not measured. Outlet concentrations were observed during each run to confirm stable conditions before the readings were taken [18].

The reduction of each measured species (SO$_{2}$, NO$_{2}$, CO$_{2}$, and CO) was quantified by comparing its inlet and outlet concentrations. Because NO and total SO$_{x}$ were not measured, the results refer to SO$_{2}$ and NO$_{2}$ only and do not represent total SO$_{x}$ or NO$_{x}$ removal. The removal efficiency was calculated as

$ \eta_i=\frac{C_{i,\mathrm{inlet}}-C_{i,\mathrm{outlet}}}{C_{i,\mathrm{inlet}}}\times100\% $

where, Ci, inlet, and Ci, outlet stand for the gas pollutant concentrations at the wet scrubber’s inlets and outlets as measured by the IoT-based multi-gas sensor node [23].

In addition to the concentration of emission gases, the temperature of the diesel engine through the silencer, wet scrubber tube, and exhaust gas temperature from the wet scrubber were also collected to determine the contribution of temperature to the pollution produced.

The wet scrubber has been successfully constructed using two different methods integrated into one system. The first system is the dipping method, which works by feeding the engine exhaust gas directly through a silencer into the water already in the wet scrubber system. The second system is the splashing or shower method, which works by spraying water into the scrubber body filled with engine exhaust gas.

This method uses a system that feeds exhaust gases (CO$_{2}$, SO$_{2}$, NO$_{2}$, and CO) from the engine directly into a volume of water. Comparison of exhaust gas concentration values based on engine-speed setting before and after wet scrubber filtration can be seen in Figure 3.

Wet gas scrubbing systems, also known as wet scrubbers, have long been regarded as a key technology for reducing industrial exhaust gas emissions. The dipping method offers a unique configuration in which the gas phase is dispersed into a continuous liquid phase [33]. The primary mechanism of this system is direct contact between the pollutant laden exhaust gas stream and the absorbent liquid, typically water or a specialized chemical solution. The absorption process allows contaminants to transfer from the gas phase to the liquid phase.

SO$_{2}$ has the highest efficiency and the best absorption efficiency among all gases, reaching an average of 78.7\% at 25\% of rated engine speed and decreasing to around 64--66\% at higher engine speeds. With removal efficiency usually surpassing 95\%, wet scrubbing is an essential technology for lowering sulfur dioxide (SO$_{2}$) emissions in industrial and marine applications [23-28]. In the dipping mode, exhaust gas is bubbled into the scrubbing liquid through a submerged inlet, as in a bubble column or jet bubbling reactor [34]. The immersion depth is the vertical distance between the gas outlet of this submerged pipe and the liquid surface; it is a geometric parameter that sets the bubble rise path and residence time and should be distinguished from the absorption mechanism itself. Compared to other exhaust gas components, sulfur dioxide is highly soluble in water. When SO$_{2}$ exhaust gas comes into contact with the water in the scrubber, the SO$_{2}$ gas rapidly transfers from the gas phase to the liquid phase. Deeper immersion increases the residence time of gas bubbles in the liquid, allowing for more thorough absorption of SO$_{2}$ into the aqueous phase [34-35]. Engine load and speed also have a big effect on how well lowering SO$_{2}$ exhaust gas concentration works. The more power the engine has to make, the more fuel it needs. As a result, the total amount of SO$_{2}$ released per hour goes up in a straight line as the load goes up. In general, raising the engine speed (particularly when the load stays the same or goes up) raises the fuel mass flow rate, which in turn raises the amount of SO$_{2}$ in the exhaust gas.

CO removal is relatively high and decreases monotonically with engine speed, from 71.1\% at 25\% of rated speed to 57.3\% at 100\% of rated speed ( Table 2). Higher engine speeds typically result in increased exhaust gas flow rates and temperatures, which can reduce the residence time of the gas within the scrubber and potentially lower the CO capture efficiency unless the solvent concentration or contact area is adjusted accordingly. This downward trend is reflected in the narrowing inlet--outlet gap in the grouped bar chart of Figure 3d and in Table 2: the higher the engine speed, the lower the efficiency. The inlet CO concentration (about 390--1,000 ppm) is the second highest of the measured gases after CO$_{2}$ (about 1,200--4,000 ppm). Compared to other pollutants such as sulfur dioxide (SO$_{2}$) or ammonia (NH$_{3}$), carbon monoxide (CO) is less soluble in pure water. Therefore, the efficiency of CO removal in a bubble scrubber depends heavily on the interfacial contact area and an optimal mass transfer coefficient [36]. The CO removal efficiency is relatively high, ranging from 57.3\% to 71.1\%; this is because reduction efficiency is strongly influenced by physical parameters such as bubble diameter (which is determined by the sparger design), liquid column height, and gas surface velocity. Controlling bubble size through surface tension stabilization and preventing coalescence is key to maximizing the mass transfer surface area. Although pure water alone is not sufficient to reduce CO, the incorporation of advanced oxidation systems (such as the Fenton process) or the use of copper-based absorbents (CuCl) in balloon column reactors can significantly improve CO reduction efficiency, often exceeding 80\% in practical applications such as waste incinerators [37].

CO$_{2}$ removal is low and inconsistent (mean 5.6–9.8\%; Table 2) because CO$_{2}$ is difficult to absorb by conventional water-based wet scrubbers. The 25\% speed setting provides the highest mean efficiency (9.8\%), while the 100\% setting provides the lowest (5.6\%). Lower engine speeds, on the other hand, prolong the gas residence time, which may improve absorption efficiency. However, the decreased flow momentum may hinder the gas-liquid mixing intensity necessary for ideal mass transfer [38]. The temperature and flow rate of exhaust gas are changed by engine speed, which affects the scrubber’s residence time [18]. Greater engine speeds enhance flow momentum but shorten the time bubbles have to react with the solvent, which could affect efficiency if the system isn’t tuned for these higher loads [38]. Standard wet scrubbers typically have very little effect on removing CO$_{2}$ from ship exhaust. This is mostly because CO$_{2}$ is poorly soluble in the aqueous solutions, like seawater, that are commonly used in these scrubbers. Less than 10\% CO$_{2}$ absorption efficiency has been demonstrated by wet scrubbing methods using specific urea solutions. Conventional wet scrubbers (which use water or weak alkaline solutions such as NaOH) have difficulty efficiently removing CO$_{2}$ due to its low solubility in water. Unlike sulfur dioxide, which is highly acidic and easily absorbed by alkaline solutions, CO$_{2}$ is a weak acid and does not readily dissolve or react with water or weak alkaline absorbents.

Figure 3c shows that the inlet--outlet difference for NO$_{2}$ is relatively small. Very low and variable efficiency NO$_{2}$ shows the lowest efficiency (2--5\%), with many values of 0\% (meaning no absorption at all in some measurements). This phenomenon occurs due to the low solubility of NO$_{2}$ in water. Although NO$_{2}$ is about five times more soluble in the aqueous phase than NO, its solubility is still insufficient to allow for significant absorption and reduction of NO$_{x}$ from exhaust gases. Furthermore, NO is the primary component of NO$_{x}$ and has low solubility in aqueous solutions, making it difficult to remove from exhaust gases through scrubbing alone [39]. Since NO can only be partially reduced under atmospheric conditions, pure water without chemical agents can absorb NO$_{2}$ with very little efficiency. Even after NO is oxidized to NO$_{2}$, plain water absorbs NO$_{2}$ only slowly, so a subsequent reduction stage with agents such as sulfite or thiosulfate is typically required [40]. The present NO$_{2}$ results are consistent with literature showing that pure-water wet scrubbing has only a limited capability for NO$_{x}$ control, particularly when NO dominates the NO$_{x}$ fraction; meaningful NO$_{x}$ removal generally requires oxidants, alkaline absorbents, or other supporting methods. Because NO was not measured in this study, total NO$_{x}$ removal could not be determined. Reported approaches for enhancing NO$_{x}$ removal include reducing agents [41], an oxidation stage followed by sulfite or thiosulfate reduction [40], NaClO$_{2}$-based oxidative absorption [42], ozone pre-oxidation combined with wet absorption [43], and non-thermal plasma with catalytic oxidation and a wet electrostatic precipitator [44]; these options are compared in recent reviews [23-45]. Their reported efficiencies depend strongly on absorbent concentration, pH, and gas composition and are therefore not directly comparable with the plain-water results of the present prototype.

The wet scrubber shower method uses spray nozzles that break the liquid into millions of fine droplets (atomization). This creates a very large surface area of contact between the exhaust gas and the liquid. The inlet and outlet exhaust gas concentrations under this shower method at different engine speeds are shown in Figure 4. The smaller the size of the water droplets, the greater the surface area to volume ratio, which significantly increases the rate of exhaust gas absorption. The system must be designed efficiently to balance high removal rates with economical pressure drop, which is generally recommended to be between 15 and 50 mmH2O/m for the packing material to avoid a decline in machine performance [46-47].

Optimal CO$_{2}$ removal efficiency is achieved at 75\% of rated engine speed. Increasing the speed to 100\% actually reduces efficiency significantly, likely due to excessive droplet atomization, which reduces the gas liquid contact time. The shower method in a wet scrubber works by dispersing the cleaning liquid through a nozzle to create fine droplets that fall through the gas stream. The CO$_{2}$ reduction efficiency of the wet scrubber system developed was only able to reduce the CO$_{2}$ emission concentration produced by the diesel engine by 10\%. This is because the shower method tends to have moderate gas absorption efficiency compared to the bubble column or dipping methods [48]. Smaller particles increase the surface to volume ratio, thereby enhancing absorption capacity. However, if the particles are too small, the momentum of the gas flow can carry them out of the system, potentially reducing efficiency [49].

SO$_{2}$ showed the highest mean efficiency at 50\% of rated engine speed, but with high variability at that setting, indicating less consistent results. Overall, SO$_{2}$ has the highest efficiency of the measured gases, with mean values of 61.5–70.4\% across the four speed settings (individual readings ranging from 60–79\%, Table 3). This method works by spraying an absorbent liquid (usually water or an alkaline mixture such as lime) through a nozzle into the exhaust gas stream. When water droplets (spray) come into contact with the gas, mass transfer occurs, causing SO$_{2}$ to dissolve into the liquid droplets. High efficiency is achieved due to the small droplet size of the water stream; the smaller the water droplet size, the greater the total contact surface area between the liquid and the gas, thereby maximizing SO$_{2}$ absorption. Additionally, the higher the volume of liquid sprayed relative to the volume of passing gas, the higher the absorption efficiency [50]. The efficiency of SO$_{2}$ removal can be worsened by increasing the exhaust gas flow rate, which is frequently brought on by greater engine speeds. Even though the scrubber entrance is where most of the scouring takes place, a large increase in flow rate can cut the residence period in half, which could leave some of the gas untreated if the L/G ratio is not changed [18]. Spray hydrodynamics, liquid flow rates, and the tower’s physical dimensions all have a significant impact on the effectiveness of SO$_{2}$ removal. The accessible surface area for absorption is determined by the droplet’s velocity and distribution.

NO$_{2}$ is removed appreciably only at low engine speed (25\% of rated speed). An increase in engine speed results in a decrease in efficiency and significant variability, indicating that NO$_{2}$ is less soluble in water and sensitive to shower flow turbulence. A shower based scrubber’s effectiveness starts with the amount of NO$_{2}$ that enters the system, which is mostly determined by engine speed. The combustion chamber has a higher filling efficiency and a higher oxygen concentration at lower engine speeds, as reported for a different diesel engine at about 1,200 rpm [51].

In that engine, increasing the speed toward 3,200 rpm, which is above the 2,200 rpm rated speed of the engine used here, reduced the available oxygen and reaction time and hence the raw NO$_{x}$ output; the same qualitative mechanism may contribute to the speed dependence observed in the present tests, although engine-out NO$_{x}$ formation was not characterized in this study. The shower method frequently needs chemical or physical support to accomplish significant NO$_{2}$ reduction. For instance, NO$_{x}$ removal efficiency of roughly 66\% have been attained by systems that use ultra-low frequency wave treatment of the water [52]. Electrolyzed water is another method that can achieve reduction efficiencies of 43.2\% to 48.9\% [7].

CO shows the highest efficiency at 25\% of rated speed (72.3\%), but the trend with engine speed is not monotonic: efficiency falls to 54.5\% at 50\%, rises to 62.5\% at 75\%, and drops to 41.1\% at 100\% of rated speed ( Table 3). Overall, lower engine speeds favour CO removal, consistent with the longer gas--liquid contact time at lower exhaust flow, while the intermediate rise at 75\% indicates that factors other than flow, such as spray--gas mixing and inlet concentration, also play a role. Spray towers offer great collection efficiency for coarse PM because their primary method of particle collection is impaction. For particles larger than 5 $\mu$m, spray towers often have clearance rates of up to 90\%. Particles with a diameter of 3 to 5 $\mu$m have removal efficiency ranging from 60 to 80\%. Removal efficiencies drop to less than 50\% below 3 $\mu$m. Applications for spray towers include dust management in fertilizer factories, pigment activities, and PM emissions from grinding operations.

Combining the dipping/bubble method with water spraying can significantly improve cleaning efficiency compared to using just one method. The inlet–outlet gas concentrations and removal efficiencies for this hybrid scheme are presented in Figure 5 and Table 4 respectively. In combined spray-and-bubble scrubbers reported in the literature, much of the pollutant load is removed in the first contact zone, while the second zone provides additional capacity for the remaining gas [48]. The very high efficiencies reported for such systems were obtained for specific pollutants, absorbents, and operating conditions; they are not results of the present prototype and cannot be attributed to the contact geometry alone.

CO$_{2}$ efficiency remains relatively stable in the range of 8--11\% across all engine-speed settings, with the differences between settings not particularly significant. SO$_{2}$ gas exhibits the highest efficiency, exceeding 75\% at all speed settings; notably, 25\% of rated speed already provides very good efficiency (79.5\%), indicating that the scrubber is highly effective for SO$_{2}$ even at low speeds. NO$_{2}$ efficiency is the lowest among the four gases, generally below 10\%, and engine speed has almost no systematic effect on NO$_{2}$ reduction, so additional treatment methods are needed for this gas. CO efficiency is highest at 25\% of rated speed (78\%) and then decreases at higher speeds, most likely because of shorter gas--liquid contact time.

The hybrid configuration that combines spray contact with a dipping/immersion zone can improve gas-liquid interaction because the spray section enlarges interfacial area through droplet formation, while the dipping section extends gas residence time and promotes secondary absorption. This explanation is consistent with a combined spray-and-bubble wet scrubbing concept, where simultaneous SO$_{2}$ and NO$_{x}$ removal increased relative to single spray or single bubble operation because gas first undergoes removal in the spray zone and then further removal in the bubble/immersion zone. In that research, the combined system improved NO$_{x}$ removal by up to 17--18\% and SO$_{2}$ removal by 11--13\% compared with separate units, indicating that hybrid contact patterns can strengthen pollutant capture [34]. Two differences limit the transferability of these gains. First, the flow path in the present prototype is reversed: the exhaust is first bubbled through the submerged inlet and only then passes the demister and the spray zone ( Figure 2). Second, the cited system used ozone pre-oxidation, so its performance gain cannot be attributed to the contact geometry alone or transferred directly to plain-water operation.

The removal efficiency SO$_{2}$ is usually the highest among the tested gases because SO$_{2}$ is an acid gas that reacts quickly in alkaline liquid medium and has a relatively high solubility, it typically has the best removal efficiency of all the gases examined. Since SO$_{2}$ can be transformed into sulfite and bisulfite species after dissolving, absorption is regulated by both liquid-phase reactivity and physical dissolution. For this reason, wet scrubbers particularly spray towers are universally acknowledged as being highly successful in controlling SO$_{2}$. According to research on spray towers, the liquid-to-gas ratio, alkalinity/pH, inlet SO$_{2}$ concentration, and gas temperature all have a significant impact on SO$_{2}$ removal. According to a reviewed spray-tower study referenced in the literature, SO$_{2}$ removal might achieve 100\% at inlet values of 500–1500 ppmv at an L/G ratio of roughly 3 L/m\textsuperscript{3} and pH 11.7. Thus, the hybrid dipping–spray scrubber exhibiting the highest $\eta$ for SO$_{2}$ is consistent with earlier research on wet scrubbers. Nevertheless, the hybrid efficiency of 75–79\% obtained here remains clearly below the 95–98\% values reported for full-scale marine and industrial scrubbers. We hypothesize that four prototype-specific factors contribute to this gap, although they were not tested individually in this study: (i) the scrubbing liquid was plain fresh water with no deliberate alkalinity, so absorption relied on physical dissolution rather than acid--base reaction; (ii) the compact drum geometry limits gas residence time compared with tall spray towers; (iii) the liquid-to-gas (L/G) ratio delivered by the single shower circuit, which could not be determined because neither the liquid nor the exhaust-gas flow rate was measured, may be lower than in high-efficiency spray towers; and (iv) the raw, uncooled exhaust enters the scrubber at high temperature, which reduces gas solubility. If confirmed, each of these factors could be addressed—through alkaline dosing, increased liquid level and spray flow, or pre-cooling—which would offer a route toward higher SO$_{2}$ capture in future versions of the device [53].

Although it can still be decreased in a wet scrubber under the right chemical conditions, NO$_{2}$ performance is typically lower than SO$_{2}$. The primary cause is that wet scrubbing nitrogen oxides is more difficult than wet scrubbing sulfur dioxide. Since NO has a low water solubility and dominates NO$_{x}$ in many exhaust streams, efficient removal frequently necessitates previous oxidation to NO$_{2}$ or other more soluble forms. Oxidation potential and pH have a significant impact on the chemistry even in the presence of NO$_{2}$. A comparison of wet scrubbing for SO$_{2}$ and NO$_{x}$ revealed that while effective NO$_{x}$ capture required an oxidizing absorbent and an ideal pH of about 8 to balance oxidation and absorption, SO$_{2}$ removal was comparatively simple [23].

Because every pollutant absorbs differently in liquid media, the hybrid wet scrubber that combines dipping and spray techniques demonstrated varying gas reduction efficiencies. Wet scrubbing is one of the most efficient methods for removing sulfur because SO$_{2}$ is extremely soluble and reacts quickly in alkaline solution. Higher liquid to gas ratios and alkaline circumstances can result in almost total SO$_{2}$ removal, according to spray tower research [53]. Although hybrid gas-liquid contact can still enhance performance by extending residence time and interfacial area, NO$_{2}$ removal is more challenging because nitrogen oxide absorption is highly dependent on oxidation-reaction routes and pH equilibrium. Because physical dissolution alone is weak, CO$_{2}$ reduction in wet scrubbers is typically limited unless the absorbent is strongly alkaline; literature reports that meaningful CO$_{2}$ capture requires high pH and reactive additives like Ca(OH)$_{2}$; CO, although removed comparatively well in this study thanks to the large interfacial area and extended residence time of the hybrid contact pattern, is intrinsically difficult to scrub because conventional wet scrubbers are not designed for carbon monoxide capture; industrial CO separation typically relies on adsorption, membranes, cryogenic systems, or specialized chemical absorbents rather than water-based scrubbing, which explains why CO efficiency in this study fell sharply once exhaust flow increased [19]. Consequently, the basic physicochemical characteristics of the gases and earlier journal articles on wet scrubber systems are compatible with an efficiency trend observed here, namely SO$_{2}$ > CO > CO$_{2}$ > NO$_{2}$.

Table 5 summarizes the mean removal efficiency ($\eta$, $\pm$ standard deviation from five replicate readings) of the four target gases under three scrubber configurations across four engine-speed conditions. The table also lists the percentage-point (pp) efficiency gain $\Delta$ of the hybrid configuration relative to the better-performing single mode; $\Delta$ is computed from the reported rounded mean values. Averaged over all four engine-speed settings, the hybrid mode improves SO$_{2}$ removal by about 6.1 percentage points over dipping-only and 10.4 percentage points over shower-only operation, and improves CO removal by about 3.5 and 10.4 percentage points, respectively. The gains for CO$_{2}$ (1.6–1.9 pp) and NO$_{2}$ (1.1–1.3 pp) are small in absolute terms, consistent with the solubility-limited behavior of these gases in pure water. The hybrid mode was, however, not superior at every individual condition: at 100\% of rated speed its CO removal (49.0\%) was lower than that of the dipping mode (57.3\%), at 75\% of rated speed its CO$_{2}$ removal (8.0\%) was lower than that of the shower mode (10.1\%), and for NO$_{2}$ the differences at 25\%, 75\%, and 100\% of rated speed were negative but smaller than the standard deviations. Its advantage should therefore be read as a pollutant-specific average advantage rather than a consistent superiority.

The hybrid mode is also the most stable configuration for SO$_{2}$: its coefficient of variation across consecutive readings remains between 1.1\% and 6.5\% at every speed setting, and its mean efficiency varies only from 75.2\% to 79.5\% across the whole operating range, whereas the dipping mode falls from 78.7\% to about 64\% and the shower mode fluctuates between 61.5\% and 70.4\%. This indicates that combining spray contact (large interfacial area) with bubble immersion (long residence time) buffers the system against changes in exhaust flow.

The decline in CO removal at higher engine speeds supports the interpretation that the loss of efficiency is flow-related. As engine speed rises from 25\% to 100\% of rated speed, the inlet CO concentration roughly doubles (about 430 to 970 ppm), reflecting the higher exhaust mass flow; over the same range, CO efficiency drops from 72.3\% to 41.1\% in shower-only mode but only from 77.8\% to 49.0\% in hybrid mode, and the immersion zone of the hybrid system partially compensates for the shorter droplet contact time. Because CO is only weakly soluble in water, its capture is especially sensitive to residence time, which explains why the flow effect is strongest for this gas.

From an environmental-management perspective, the relevance of this prototype lies in its accessibility. Because the device is built largely from recycled industrial oil-storage drums and operates with plain fresh water, it represents a low-capital, low-complexity auxiliary exhaust gas cleaning approach that could realistically be adapted for small vessels, port auxiliary equipment, harbor workboats, or stationary diesel generators in low-resource settings, where certified marine scrubbers or low-sulfur fuels may be economically out of reach. In such contexts, even partial abatement of SO$_{2}$ (about 75--79\%) and CO (up to 78\%) constitutes a meaningful local air-quality benefit.

The use of recycled oil drums also embodies a circular-use principle, extending the service life of industrial waste material as pollution-control equipment. However, this choice is appropriate only at the prototype and demonstration stage. Real shipboard exhaust gas cleaning systems cannot rely on non-standard containers: they must satisfy class and flag-state requirements for corrosion resistance against acidic condensate, pressure and vibration resistance, fire safety, maintainability, and back-pressure limits on the engine. Scaling this concept toward marine service would therefore require certified materials and a formal engineering assessment, while the recycled-drum approach remains valid for land-based or experimental applications.

Finally, wet scrubbing does not destroy pollutants; it transfers them from the gas phase to the liquid phase. Wash-water composition and pH were not measured in this study. Based on scrubbing chemistry and the literature on scrubber effluents, however, the recirculated water would be expected to accumulate dissolved SO$_{2}$ (as sulfurous acid, sulfite, and sulfate), dissolved CO$_{2}$, and nitrite/nitrate from absorbed NO$_{2}$, and therefore to become progressively more acidic. Combustion-derived particles may also be retained, but particulate collection is a separate mechanism from dissolved-gas absorption and was not quantified here. If discharged untreated, especially in ports or estuaries, such effluent could create secondary water pollution of the kind already documented for open-loop marine scrubber discharge [11-15]. Responsible deployment of the proposed system requires closed-loop operation with periodic water replacement, neutralization (for example with lime or caustic dosing), sedimentation or filtration of particulates, and compliant disposal of the residue. Quantifying wash-water pH, sulfate, nitrate, and suspended-solid loads is identified as a priority for future work, so that the trade-off between air-pollution reduction and water-environment risk can be evaluated explicitly.

This study set out to determine whether a structural combination of bubble-immersion (dipping) and spray (shower) gas--liquid contact can outperform either contact mode alone for cleaning marine diesel exhaust. The prototype, built largely from recycled oil drums and tested at 25--100\% of rated engine speed, answered this largely affirmatively: averaged over the four engine speeds, the hybrid configuration delivered the highest removal efficiencies of the three modes for all four gases and the most stable SO$_{2}$ removal, gaining on average about 6--10 percentage points for SO$_{2}$ and 3--10 percentage points for CO over the single-mode systems. The advantage was not universal, however: at 100\% of rated speed the dipping mode removed more CO (57.3\% vs 49.0\%), and at 75\% of rated speed the shower mode removed more CO$_{2}$ (10.1\% vs 8.0\%).

SO$_{2}$ and CO are the pollutants best suited to the current device. Hybrid-mode SO$_{2}$ removal remained at 75--79\% across the entire speed range with a coefficient of variation below 7\%, and CO removal reached 78\% at low engine speed, although it declined with increasing exhaust flow because of reduced gas--liquid contact time.

CO$_{2}$ and NO$_{2}$ remain the weak points of the system. With pure water as the scrubbing liquid, CO$_{2}$ removal was limited to 8--11\% and NO$_{2}$ removal stayed below 10\%, reflecting the low aqueous solubility of CO$_{2}$ and the dominance of poorly soluble NO in the NO$_{x}$ fraction. The device should therefore be regarded as a transitional, auxiliary emission-control measure for SO$_{2}$ and partial CO abatement, not as a carbon-capture or NO$_{x}$-control technology; meeting those needs would require alkaline absorbents, oxidants, or a multi-stage treatment system.

The main limitations of this work are its single experimental campaign on one small engine, the use of fresh water only, the unquantified wash-water quality, and a prototype structure that does not yet meet marine equipment standards. Future work should test seawater and alkaline or oxidant-enhanced scrubbing solutions, add independently repeated experiments with full uncertainty analysis, measure wash-water pH and pollutant loads to manage the air-to-water pollution transfer, and evaluate long-term running stability, corrosion, and engine back-pressure before any practical deployment.