2. NRMM Emission Studies
The study of NRMM emissions in the literature has largely been overshadowed by their on-road counter parts due to data on NRMM not being as freely available
[4][9]. It was assumed until the 1990s that emissions from non-road sources were not as significant as those from on-road sources, but studies in the early 1990s suggested otherwise
[4][5][9,10]. Puranen and Mattila
[4][9], for example, discovered in a study in 1990 that the total amount of diesel fuel consumed by work machines was 780,000 m
3, which equated to about 30% of all traffic consumption in Finland. In comparison the amount of gasoline consumed by them was only 69,000 m
3, equating to 3% of all traffic consumption. They also found that the NO
x emissions from work machines accounted for 15% of total NO
x emissions, and that PM, CO, CO
2 and HC emissions from the work machines accounted for between 4% and 10% of total such emissions.
As a consequence of this oversight and lag in emission standards, the concern about emissions from NRMM has risen considerably in recent decades as their relative contribution to anthropogenic emissions continues to increase while mitigation efforts for on-road vehicle emissions improve, leading to a decrease in their emissions. This means that NRMM will eventually surpass on-road vehicles as the leading source of mobile pollution
[6][11]. This rise in the share of emissions for NRMM was predicted over two decades ago, yet fewer studies regarding emissions of NRMM sources have been conducted in the past twenty years compared to the number of studies conducted in relation to emissions from on-road vehicles
[4][5][6][7][9,10,11,12].
Despite increased interest, many studies have highlighted the challenges faced when investigating NRMM emissions, the first being that there is still a considerable lack of off-road emission data available, particularly in-use emission data
[8][13]. This lack of data and related statistics also leads to inaccuracies in estimating overall emissions and inventories from NRMM. Differences in engine design, operating conditions, engine load and fuel use result in varying exhaust compositions and amounts, making predictions of emissions difficult if data are limited
[9][14]. Generally, emission tests were carried out in laboratory settings mimicking the in-use operation of NRMM engines. However, such tests have been shown to be inaccurate when compared to real-life emissions
[10][15], a scenario highlighted by the 2015 Volkswagen scandal
[11][16]. This brings to light the fact that real-life emissions are different from emissions created in laboratory settings. Laboratory testing often fails to accurately account for real-life factors such as the duration and intensity of tasks performed by an NRMM and the different levels of activity within that operation (e.g., idle, moving, etc.), as generally only the engine/chassis is tested
[10][15]. To overcome the uncertainties in laboratory testing, legislators have called for the measurement of Real Driving Emissions (RDE). Unfortunately, collecting data from NRMM while in operation (real conditions) also encounters several challenges. Past studies have highlighted difficulties in quantifying exhaust emissions in comparison to on-road machinery due to the wide range of activities undertaken by NRMM and the operating cycles employed
[12][13][17,18], as well as difficulties in installing the required instrumentation
[14][19], and the general high cost associated with such tests. However, reliable results that cannot be replicated in laboratory settings are generated. For this reason, studies have moved towards the use of Portable Emission Measurement Systems (PEMS) and similar devices to obtain in-use exhaust emissions of NRMM
[15][16][20,21]. Considerable work has been carried out in recent years using PEMS on a range of different NRMM
[15][16][20,21], including several types of construction and agricultural machinery
[6][7][8][12][17][18][19][20][21][22][23][24][11,12,13,17,22,23,24,25,26,27,28,29].
Although fewer in number than on-road machinery, NRMM have been shown to have considerable impact on the production and emission of NO
x and PM. Recent studies from China illustrated that NRMM produced the same concentrations of NO
x and PM as was produced by all on-road sources in 2018, which had a notable impact on air quality in the form of smog
[7][25][12,30]. Many papers have also investigated the health impacts
[26][31] and chemical composition of such emissions
[25][27][28][30,32,33]. However, due to the diverse nature of NRMM, past literature has generally focussed on assessing the emissions of a few types of machinery within sectors of interest. Studies regarding emissions from construction and agricultural machinery are well documented, with additional studies also covering emissions from forestry, cargo, port and handheld and domestic gardening equipment.
2.1. NRMM Emissions in Agriculture
Agriculture represents one of the oldest and most important global industries, and it is responsible for approximately 5% of global energy consumption
[29][34]. Nowadays, a variety of NRMM are regularly employed for a range of agricultural tasks. The weighted contribution of agricultural NRMM to anthropogenic emissions has been thoroughly explored in past literature. A study from Poland compared the fuel consumption of farming machinery between 2012 and 2013, and suggested that emissions from agricultural NRMM had the biggest impact on the total emission of CO, NO
x, hydrocarbons and PM
[28][33]. Similar trends were also observed throughout Europe and elsewhere
[4][7][30][31][9,12,35,36]. Emissions of NRMM from agricultural sources in China have received much attention in the past decade due to their impact on the air quality, with a series of studies concluding that agricultural machinery is an important NRMM source that contributes significantly to these emissions
[7][30][12,35].
Additionally, an investigation by the Swiss Federal Office for the Environment into the fuel consumption and emissions of NRMM found that agricultural machinery alone accounted for 300 tonnes of PM in 2010. Interestingly, this was more than four times higher than the PM produced by construction machinery, even though diesel consumption was nearly 20% lower for agricultural machinery. This was attributed to the long lifespan of agricultural NRMM and the lack of retrofitted PM filters
[32][37]. Therefore, it is estimated that agricultural NRMM emissions will not reduce to the same, or greater, extent as the emissions of construction NRMM that already have such measures in place. Fortunately, a reduction in agricultural emissions was determined for the 1985–2050 period (the Swiss report covers the period from 1980 and includes forecasts up to 2050, with 2010 serving as reference year), with similar trends also observed by Winther and Nielsen
[21][26] and Hou et al.
[33][38]. This reduction is particularly true for NO
x and PM emissions due to improved diesel fuel standards, emission regulations and the proposed use of exhaust filters
[33][38].
Improved emission testing methods for agricultural NRMM constitute another topic that has been well documented in recent times. Several studies have investigated the deviations between exhaust emissions of NRMM in real-world and laboratory settings. Pirjola et al. (2017)
[34][39] compared exhaust emissions emitted by a tractor in real-world conditions to those of a similar engine using a dynamometer. It was found that NO
x emission factors were approximately 50% higher in real-world conditions compared to laboratory tests. Deviations were also observed for the PM emissions with the production of nucleation mode particles in real-world conditions that were absent in laboratory testing, because of the absence of pollutants, e.g., ammonia, in the lab. With the introduction of RDE measurement, many studies have assessed the suitability of RDE devices for measuring the emissions of agricultural NRMM. PEMS are the most widely assessed RDE devices and have been included in many NRMM emission studies
[17][21][23][24][22,26,28,29]. Szymlet et al. (2018)
[23][28] even compared the emissions of a passenger car to a tractor using PEMS, illustrating emissions from NRMM to be several times higher than those from the on-road source.
Tractors are the most studied of the agricultural NRMM due to them being the most polluting of farming machinery
[35][40]. Hou et al. (2019)
[22][27] investigated the emission characteristics of 22 different types of agricultural NRMM including tractors, harvesters and micro-tillers and determined that tractors accounted for over 80% of agricultural NRMM emissions in Beijing in 2016. Many papers have focussed on assessing the engine performance of tractors and emissions during different operating modes in an effort to optimise the engine performance
[24][35][36][37][38][29,40,41,42,43]. As a result, several simplified emission testing methods have also been developed for tractors. Janulevičius et al., 2013
[35][40], determined the suitability of measuring exhaust emissions during operation by using data collected in ECU load profiles of the tractors ECUs. Similarly, Ettl et al.
[39][44] developed an alternative method based on torque data and ECU engine speed from long-term tractor operations. This method used a simplified test stand opposed to portable measurements to determine the real-world emissions and fuel consumption measurements of tractors.
As a result of the serious implications associated with the release of NRMM emissions in agriculture, many studies have also focussed on the mitigation and reduction of pollutants from exhaust gases. Lovarelli and Bacenetti, 2019
[40][45], describe some technological solutions suitable for agricultural tractors and self-propelled machines. Examples of these devices are Diesel Particulate Filters (DPF), Exhaust Gases Recirculation (EGR) and Selective Catalytic Reduction (SCR).
2.2. NRMM Emissions in Construction
Construction NRMM emissions have undergone many of the same studies as those for the agricultural NRMM. Both NRMM types generally utilise diesel engines, which have been shown to favour NO
x and PM production. However, unlike for agricultural NRMM, the literature on construction machinery shows that they have not seen the same decreasing trend in energy consumption. The Swiss Federal Office for the Environment carried out an investigation into the energy consumption and emission trends of NRMM for 1980–2050. The results showed that the energy consumption of construction NRMM from 1980 to 2010 almost trebled, with a further 20% increase expected by 2050
[32][37]. Notter et al.
[31][36] also determined a 5% energy consumption increase between 2000 and 2015. This increase in energy consumption comes as a result of further urbanisation and resulting expansions of construction industries. Regarding greenhouse gas emissions, Notter et al.
[31][36] determined that although agricultural machinery surpasses construction machinery in the production of NO
x and PM, construction machinery was found to contribute the most to the emission of CO
2. However, a decrease in emissions has been observed in recent years. For example, the Swiss Federal Office for the Environment
[32][37] reported that PM emissions have fallen by 28% between 2005 and 2010, with similar reductions experienced for NO
x, as a result of strengthened regulations on air pollution and use of particle filters. This reducing trend has been predicted to continue due to improvements in engine and fuel standards, more stringent air pollution legislation and the eventual replacement of older (more polluting) machinery with newer more efficient models
[31][32][33][36,37,38].
The characterisation of the emissions from construction machinery has received much attention due to its potential as an important source of air pollution, particularly in urban areas. The major urban centre of London introduced emission standards from construction machinery in 2015 by establishing a low emission zone
[41][46]. In 2016, construction was responsible for 15% of PM
2.5 emissions and 34% of PM
10 emissions in London
[12][17]. Due to the health effects associated with the inhalation of PM and the likelihood of urban exposure, several investigations have specifically examined the measurement and composition of PM emissions from construction NRMM
[6][14][25][27][11,19,30,32]. Zhang et al.
[6][11] found similarities between construction exhaust particles and those of other diesel vehicles, containing similar proportions of water-soluble ions, organic and elemental carbon and polyaromatic hydrocarbons (PAHs). Similarly, Yu et al.
[25][30] found that PM
2.5 emissions of construction NRMM were majorly composed of carbonaceous components. These carbonaceous particulates and PAH components represent a particular health risk to the human respiratory system.
Recent studies have also increasingly sought to measure exhaust emissions from construction NRMM under actual operating conditions. Desouza et al.
[12][17] measured the exhaust emissions of 30 construction machines (including generators, excavators, dumpers, rigs, cranes and telehandlers) at active construction sites in London to evaluate the real-world emissions of construction NRMM using PEMS. Guo et al.
[21][26] carried out an additional study on 50 construction machines and 37 tractors and harvesters using PEMS to determine emission factors. Similar investigations have been carried out for excavators
[8][11][19][42][13,16,24,47], wheel loaders
[18][23], forklifts
[43][48] and motor graders
[15][20]. Other real-time exhaust methods have been investigated in the literature. Muresan et al.
[44][49] and Sennoune et al.
[45][50] used similar systems to measure the exhaust emissions of earth work machines. The use of such equipment to determine real-life measures of exhaust emissions is important for the establishment of accurate NRMM emission factors and for the establishment of emission inventories. However, they have also highlighted other potential analytical applications. Desouza et al.
[12][17] were able to detect the failure of the SCR on a telehandler that gave no warning of failure, while measuring exhaust emissions.
2.3. Other NRMM Emissions
Although NRMM from agriculture and construction contribute the most to anthropogenic emissions, NRMM from several other sectors have been studied. Several studies have focussed on emissions from forestry/logging NRMM
[46][47][48][49][50][51][52][51,52,53,54,55,56,57]. Automated logging processes are more ecological than the use of more traditional chainsaw methods
[51][56]; as such, many countries operate almost exclusively mechanised logging processes
[53][54][58,59], the atmospheric emissions of which need to be accurately determined. Lijewski et al.
[54][59] assessed the fuel usage and exhaust emissions of an entire logging process including harvesters, forwarders and transport using PEMS, illustrating once more the deviations between real-life emissions and traditional homologation tests. Harvesters were shown to contribute the most to NO
x and PM emissions due to the use of diesel engines, and forwarders contributed the most to CO and HC emissions.
The contribution of cargo and port handling equipment to regional/national emissions has also been investigated
[55][56][57][60,61,62]. Generally, cargo handling equipment (CHE) is included in construction NRMM, but studies have suggested the separate management of such equipment. The first port emission inventory was constructed in the U.S. for the Port of Long Beach in 2004
[56][57][61,62]. Zhang et al.
[57][62] developed a similar method for the estimation of CHE emissions in Nanjing Longtan Container Port, highlighting their contribution to NO
x, CO and PM emissions, with container trailers being the most polluting of the equipment.
Examples of emission testing of handheld and gardening NRMM can also be found in the literature. Lijewski et al.
[54][59] used PEMS to measure the exhaust emissions from handheld generators and chainsaws. Emission testing of garden NRMM such as lawn mowers have also been conducted
[58][59][60][63,64,65]. Priest et al.
[60][65] assessed the CO, CO
2, CH
4, non-methane hydrocarbons (NMHC) and NO
x emissions from 16 in-use lawn mowers, the results of which were used to estimate the total emissions from lawn mowers in the Newcastle region of Australia. Similarly, Millo et al.
[61][66] assessed the emission characteristics of 14 different types of common NRMM engines, including engines from lawn mowers, chainsaws, trimmers and snow removal equipment. Millo et al.
[61][66] then compared these emissions with U.S. emission standards in an effort to form a basis for the European emission standards. Although most studies have focussed on industrial NRMM and measurement of their emissions, it is important to also note the contribution of more domestic handheld and gardening NRMM.