Cems Download Data Epa Continuous Emissions

Continuous Emission Monitoring System

According to CEMS measurements, for a continuous refinery combustion operation, 3.5 tons yr−1 of SO2 and 940 tons yr−1 of NOx were emitted to the ambient environment for 150 × 106 British thermal units (Btu) per hour burning of refinery fuel gas (EPA, 2011).

From: Process Safety and Environmental Protection , 2019

Environmental Monitoring

Ravi K. Jain Ph.D., P.E. , ... Jeremy K. Domen M.S. , in Environmental Impact of Mining and Mineral Processing, 2016

Particulate Matter Monitoring

The monitoring of dust emissions is important for determining the effectiveness of preventative measures and for determining the potential hazard. Emissions monitoring systems like CEMS can monitor PM indirectly, but only if emission streams are constant and the system is properly calibrated (U.S. Environmental Protection Agency (U.S. EPA), 2002). For highly variable emissions, PM can be monitored using passive or active systems. Passive systems simply measure PM deposition over a period of time and involves collecting dust using a flat surface, glass slides, stick pads, bowls, or a cylindrical container (Department of Environment Climate Change and Water NSW, 2010). Passive systems are usually exposed and measured on a weekly or monthly basis. Active systems utilize gravimetric measurements to provide a time-weighted average PM concentration and are the standard EPA method for measuring PM (U.S. Environmental Protection Agency (U.S. EPA), 2002). In an active system, a measured volume of air is drawn through a filter and the difference in the weight of the filter is used to determine PM concentration (Department of Environment Climate Change and Water NSW, 2010; U.S. Environmental Protection Agency (U.S. EPA), 2002). A combination of both monitoring systems should be used to evaluate the effectiveness of dust control measures.

Personal and mobile dust monitors are available to measure individual exposure or to attach to vehicles to quantify dust generation during operation. These dust monitors can be active systems (utilizing gravimetric measurements), or may use light scattering as a passive method to determine dust exposure.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978012804040900005X

Alberta Oil Sands

X.L. Wang , ... K.E. Percy , in Developments in Environmental Science, 2012

8.3.3 Gas and PM Concentrations and Emission Rates

The continuous measurement devices detected short-duration stack emission changes indicative of process and control device variations, as illustrated in Figure 8.7. Stack parameters and gas concentrations were relatively stable over the test periods, in contrast to real-time concentrations from heavy haulers in mining operations (Watson et al., 2010a). The PM spikes in Stack C corresponded to NO spikes, probably due to short-duration changes. Such correspondence was not found in Stack A or B samples during this study. Detection of short-term variations might be useful for explaining test-to-test differences in ERs and compositions.

Figure 8.7. Examples of real-time data for: (A) stack temperature, (B) velocity, (C) CO, (D) NO, (E) PM_2.5, and (F) PM₁₀ concentrations from one run at each stack.

Table 8.5 shows that CO₂ had the highest ERs among all pollutants, ranging from 1.8   ×   10⁵ to 2.7   ×   10⁵  kg/h. While H₂S ERs (0.003–0.038   kg/h) were low, large variations were found in NH₃ ERs. Stack C emitted 1–2 orders of magnitude lower NH₃, only 1.1% and 0.2% of levels in Stacks A and B, respectively, which explains the higher acidity of PM_2.5. ERs ranged from 500–1599   kg/h for CO, 201–   >   1050   kg/h for SO₂, 132–295   kg/h for NO _x , and 8–49 and 11–68   kg/h for PM_2.5 and PM₁₀, respectively.

Table 8.5. Average Gas and PM Concentrations (Under Standard Conditions) and Emission Rates for the Three Stacks

		Concentration (mg/m3)			Emission rates (kg/h)
Measured species		Stack A	Stack B	Stack C	Stack A	Stack B	Stack C
Gases	CO	698   ±   27	929   ±   43	451   ±   74	1599   ±   54	980   ±   44	500   ±   73
	CO2	(1.14   ±   0.02)   ×   105	(1.68   ±   0.03)   ×   105	(2.40   ±   0.03)   ×   105	(2.61   ±   0.05)   ×   105	(1.77   ±   0.02)   ×   105	(2.70   ±   0.05)   ×   105
	NO	128   ±   4	126   ±   2	167   ±   11	295   ±   11	132   ±   2	187   ±   12
	NO2	0.0   ±   1.2	0.0   ±   0.7	0.0   ±   1.3	0.0   ±   2.7	0.0   ±   0.7	0.0   ±   1.4
	NO x	128   ±   3	126   ±   2	166   ±   10	295   ±   11	132   ±   2	186   ±   11
	NH3	7.3   ±   1.1	82.0   ±   21.3	0.16   ±   0.01	16.6   ±   2.4	86.4   ±   22.9	0.18   ±   0.01
	SO2	>   461	689   ±   122	177   ±   22	>   1050	727   ±   132	201   ±   28
	H2S	0.017   ±   0.008	0.005   ±   0.001	0.003   ±   0.001	0.038   ±   0.017	0.005   ±   0.002	0.003   ±   0.001
PM	PM1	13.3   ±   1.1	5.6   ±   0.1	36.7   ±   2.7	30.6   ±   2.7	5.9   ±   0.1	41.3   ±   3.4
	PM2.5	21.5   ±   2.1	7.6   ±   0.3	38.0   ±   3.0	49.1   ±   5.0	8.0   ±   0.3	42.7   ±   3.8
	PM10	29.6   ±   3.5	10.1   ±   1.2	38.2   ±   3.2	68.0   ±   8.4	10.7   ±   1.4	43.0   ±   4.0
	PM25	29.6   ±   3.5	10.3   ±   1.4	38.2   ±   3.2	68.1   ±   8.4	10.9   ±   1.6	43.1   ±   4.0

Table 8.6 shows that for Stack A, the SO₂ ER from dilution tests was one-eighth that from CEMS during the same sampling period because the potassium carbonate (K₂CO₃) on the backup filter was completely consumed by SO₂. SO₂ from CEMS and a 2007 compliance test differed <   2%. NO _x and TSP ERs from dilution tests were 52% and 17%, respectively, of ERs from 2007 compliance tests. ERs from dilution sampling (except the unquantified SO₂), CEMS, and compliance were well within Alberta's emission guidelines for these species. For Stack B, SO₂ from dilution tests was ∼   25% lower than the CEMS but was similar to the 2007 compliance tests. NO _x by dilution sampling was 45% higher than that from the 2007 compliance test. The TSP by dilution sampling was 21% of the hot filter catch and 3% of the total TSP from the compliance tests. The Stack B total TSP would exceed the emission guideline value by 51%, if the impinger catch was included. For Stack C, TSP by dilution sampling was 16% lower than the 2010 compliance test result, and both NO _x and TSP were <   16% of emission guidelines. The discrepancy in TSP between dilution sampling and compliance tests might be partially attributed to losses of >   15   μm particles in dilution sampling, positive artifacts from the impinger catch, and emission variations between the 2007 and 2008 testing periods.

Table 8.6. Comparison of Emission Rates (kg/h) from Dilution Sampling, Continuous Emissions Monitoring Systems (CEMS) During the Dilution Sampling Period, Compliance Tests (Conducted in 2007 for Stacks A and B, and 2010 for Stack C), and Alberta Environment's Emission Guidelines for Each Stack

Stacks/species		NO x a	SO2	TSP-front	TSP-total b
Stack A	Dilution	452   ±   17	>   1050	NA c	68   ±   8
	CEMS	NA	8699   ±   157	NA	NA
	Compliance	870   ±   69	8843   ±   431	NA	392   ±   48
	Guideline	1500	16,400	NA	600
Stack B	Dilution	203   ±   4	727   ±   132	NA	10.9   ±   1.6
	CEMS	NA	951   ±   288	NA	NA
	Compliance	140   ±   19	741   ±   239	52   ±   10	378   ±   50
	Guideline	NA	NA	NA	250
Stack C	Dilution	284   ±   17	201   ±   28	NA	43.1   ±   4.0
	CEMS	NA	NA	NA	NA
	Compliance	NA	NA	NA	51.09
	Guideline	1800	NA	NA	340

a

Compliance NO _x was measured by Alberta Method 7A, where NO _x in the flue gas sample was oxidized to nitrate, measured by ion chromatography (IC), and reported in term of NO₂. To be comparable to the compliance tests, NO _x from dilution sampling is also reported as NO₂ by: ER (NO₂)   =   ER (NO)   ×   46/30.

b

The compliance tests measured TSP emissions based on Alberta Method 5: TSP-front: hot filter catch; TSP-total: hot filter and impinger catches. The dilution sampling TSP uses PM₂₅ by the OPC.

c

Data not available.

Table 8.7 compares particle size distributions from this study with those measured from a 2002 in-stack survey. The methods are different, and the processes are probably better controlled for the 2008 tests, so a precise correspondence is not expected. The PM_2.5 and PM₁₀ ERs from the in-stack survey were, respectively, 63% and 15% lower than those by the dilution method, while TSP was 51% higher. The in-stack hot filter did not collect particles that would nucleate and grow upon cooling and therefore underestimated fine particle concentrations. This is probably why the hot filter PM_2.5 was so much lower than the dilution sample. The in-stack TSP (>   10   μm) fraction is uncertain and subject to contamination, as it was recovered by washing the sampling probe and cyclone with acetone (U.S. EPA, 2010). However, losses for particles >   15   μm in the dilution sampling method were not accounted for, and the OPC only measures particles ≲   25   μm.

Table 8.7. Comparison of Particle Size Distributions Measured from Stack A by an In-Stack Survey Test and the Dilution Sampling in This Study

Test name a	Test date	PM2.5 ER (kg/h)	PM10 ER (kg/h)	TSP ER (kg/h)
In-stack survey	5/1/2002–5/2/2002	18   ±   2	58   ±   10	103   ±   27
This study	8/9/2008–8/11/2008	49   ±   5	68   ±   8	68   ±   8

a

The in-stack sampling followed a setup similar to the U.S. EPA Method 201A by installing both a PM₁₀ and PM_2.5 in-stack cyclones (U.S. EPA, 2010). The condensable fraction captured in the impingers was not accounted for in the in-stack test data.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080977607000081

Back to Basics

Clark W. Gellings , in Energy Efficiency, 2013

3.2 Analysis of Electricity Use in Power Plants

3.2.1 Overview

The Electric Power Research Institute (EPRI) conducted an evidence-based analysis of auxiliary or parasitic loads (internal plant usage of power) in the U.S. fossil and nuclear generation fleet to better understand such power usage. According to conventional wisdom, internal power needs are approximately 5 to 10 percent of total generation, and this usage can vary by fuel type. Power needs are known to vary across such parameters as unit age, capacity, heat rate, capacity factor, number of starts, ambient operating temperature, and cooling water temperature.

Using the commercially available Energy Velocity database, ³ the EPRI team gathered data on power generation across the U.S. fleet, for coal plants, nuclear plants, and natural gas plants. The team analyzed internal power usage (as a percentage of gross generation) across each fuel-specific fleet, and statistically (i.e., through regression analysis) related usage to key characteristics across the fleet, including unit capacity, age, heat rate, and usage (based on information such as capacity factor and number of annual starts). The team used regression techniques to help parse the internal power requirement (on average) to each key contributing characteristic.

3.2.2 Data Sources and Quality and Statistical Approach: Power Plant Electricity Usage

Gross generation measured prior to subtracting auxiliary power for internal usage, as well as parameters such as unit age, capacity, heat rate, and capacity factor, are available by generation unit (EPA). Net generation ⁴ measured essentially at the busbar, which is the amount that is supplied to the grid, is available only at the aggregate plant level (EIA). The team analyzed data for five recent years – 2005 through 2009.

Analysts first matched up all unit information to its corresponding plant so that a consistent gross generation versus net generation composite database could be derived. Data sanitation led to some culling of entries ⁵ due to inconsistencies and imperfect plant-to-unit correspondence. Inconsistencies include data not reported for both data sets or net generation larger than gross generation. Once a sanitized database was assembled, the team ran a variety of regressions to test which explanatory variables were most critical in explaining the range of internal usage variation.

3.2.3 Overview of Electricity Usage Analysis Results: Power Plants

Across the U.S. generation fleet, there are variations in internal power usage. These may be explained in part by variations in parameters that are related to duty-cycle or overall heat rate, including unit capacity, age, heat rate, and number of starts. They may also be explained by configuration-related parameters, such as the presence/absence of particular types of pollution control equipment (e.g., scrubbers and electrostatic precipitators).

The analysis did not clearly explain variations through the detailed differences in pollution control equipment. Overall, the variations across plants and units were simply too narrow, and they were mostly superseded by macro-level indicators such as age and duty cycle. However, macro-level indicators were useful for the coal and nuclear fleets, and these indicators largely conformed to conventional wisdom as suggested above.

3.2.4 Electricity Usage in Coal Power Plants

Approximately 350 plants were in the data sample after sanitizing. These data were first separated by individual coal type, but that parsing yielded no significant difference in results from the aggregate analysis. The average power usage of internal plant auxiliaries across the sample was 7.6 percent, with a standard deviation of 2.9 percent.

There is considerable scatter in these results. In examining the full data set of roughly 1750 data points, the internal power fraction ranges from as low as 4 percent to as high as 12 to 13 percent. Of the key driving variables tested, plant heat rate was the most sensitive one, but extreme variations in heat rate only seem to capture about 40 percent of that range. The rest of the variation seems to be obscure, asystematic characteristics of individual plants, data inconsistencies, and the like.

In summary, the study data indicate that auxiliary power use is not as large as once thought. Larger plants tend to run more consistently, and start and stop less frequently; therefore, they tend to use less electrical energy internally. The age of the plant does not appear as relevant as once thought based on this investigation. Although newer plants are designed to be more efficient, they are required to be fully outfitted with emission controls and often with mechanically driven cooling towers. As such, newer plants are not more efficient in their use of auxiliary power.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123978790000153

Wireless communication

B.R. Mehta , Y.J. Reddy , in Industrial Process Automation Systems, 2015

14.7 Applications of wireless instrumentation

There can be several areas of application for wireless instrumentation. Generally, a plant is divided into two areas, in terms of scope of the production. ISBL means inside battery limit, which defines the plant production area. Many of the utilities and raw materials used for processing are not always produced inside the plant. There are items such as instrument air, cooling water, and steam that are supplied from OSBL, that is, outside battery limit, of the plant. Apart from the utilities, the raw material and produced material comes from OSBL, such as tank firms and storage areas.

Major applications are as follows.

14.7.1 ISBL applications

Major wireless applications for ISBL area include:

•

Temperature profiling of columns/reactors

•

Skin temperature measurement in boilers/furnaces

•

Pressure, differential pressure monitoring

•

Inter-unit flow measurement

•

Motor bearing/winding temperature monitoring

•

Monitoring of important manual valves/bypass valves

•

PRVs/flare valves monitoring

•

Corrosion monitoring

14.7.2 OSBL applications

Major wireless applications for OSBL area include:

•

Vibration monitoring of fin–fan coolers

•

Vibration monitoring of cooling towers

•

Vibration monitoring of other second/third-tier assets such as pumps/blowers

•

Tank farm monitoring

•

Effluent Treatment Plants/ Reverse Osmosis (ETP/RO) plants – pressure/temperature/level monitoring

•

Sea water intake facility

•

Pressure, temperature monitoring

•

Steam water analysis system

•

Continuous emission monitoring system

•

Motor bearing/winding temperature monitoring

•

Monitoring of important manual valves/bypass valves

•

Corrosion monitoring

14.7.3 Storage and tank firm monitoring

Every industrial plant needs raw material for production and they use these raw materials to produce their final output. Generally, these raw materials and final outputs are stored in a distant location from the manufacturing location. These storage tanks have different materials that need to be kept in certain conditions. For safe storage of these materials, certain temperature or pressure has to be maintained. The operator might also be interested to know the level in the tanks. So there are enough process parameters that are to be monitored or controlled. Wireless becomes an ideal solution for bringing these process parameters in the control room because of the cost advantage. Moreover, wireless gives flexibility of adding any process point without bothering about extra cable.

14.7.4 Mounting on rotating equipment

There are many equipment and mechanical instruments that are rotating in nature apart from standard pump, motor, turbine, or compressor. The low-speed rotating equipment such as dryer or kilns are huge in size and some of the measurements need to be done on the rotating body itself. Wireless instrumentation solves the age-old wear-and-tear-related issue of conventional wired device mounted on these equipment.

14.7.5 Vibration-prone area

Process variable measurements are also necessary in vibration-prone areas of reciprocating pumps, centrifuge, pneumatic conveying, and so on. The nature of process is such that these areas or systems experience lot of vibrations and huge pressures that are irregular and fluctuating. This makes instrumentation extremely challenging because conventional wired transmitters with all the terminals and junctions become vulnerable to lose connection. Wireless transmitters are best suited for these purposes.

14.7.6 Across a river

Wireless instrumentation is a boon for data transfer across a barrier where there is a difficulty of cable. Generally, two infrastructure nodes are put across two shores to enable data communication. The sensors data from the other shores can be transported over infrastructure nodes to bridge the gap between the two shores and data can be sent to the control room.

14.7.7 Moveable platform

Moveable equipment such as a trolley in a rail, cranes, or earth movers are slowly moving toward wireless instruments because of their flexibility and low maintenance. The cables of wired instruments need to be protected with armored shields because of the harsh environment they are deployed. Even protecting shield is not sufficient enough to avoid regular wear and tear. Wireless instrumentation gives the answer for all these problems posed by the harsh environment.

14.7.8 Offshore gas and oil exploration

Offshore oil and gas platforms are another segment where wireless instruments are getting lot of popularity. The reason behind this choice is slightly different than all others. Offshore platforms have very stringent power usage guidelines. As it is floating in the middle of the ocean, all the power requirements have to be met by its own generation units. Addition of any equipment or instrument is carefully evaluated in terms of its power consumption specifications. Wireless devices clearly come out a winner here, as most of the instruments are battery powered. These batteries are long lasting because of the highly advanced algorithm and sleep features that are implemented in these transmitters. As an example, for a 30-sec reporting rate transmitter, the battery can last up to 10 years for some of the vendors. For 1-sec reporting rate, the battery life can be a minimum of 2 years. Regardless to say that any wiring-related troubleshooting becomes redundant because staff and labor availability is always a question in this environment.

14.7.9 On board of a ship

Wireless instrumentation is also fast gaining popularity in ships and ocean liners. The popularity is mainly because of its simplicity and less documentation. Wireless becomes ideal for small-scale deployment. It becomes also easy for troubleshooting. Any replacement or new additions become very easy.

14.7.10 Mobility with handheld station

The popular mode of wireless system deployment in larger plants comes with wireless infrastructure development. IEEE 802.11a/b/g/n standards are the popular choice of any wireless back haul or infrastructure. This is an industry-wide common and open standard. 802.11 b/g (also known as WiFi) is used in handhelds, smart phones, laptops, and so on. Typical client-side access has WiFi capability. 802.11a/n is used for main backhaul or mesh throughput. The advanced infrastructure nodes generally have this capability of acting as WiFi access point and also as meshing backhaul along with industrial protocol. A single infrastructure supports both WiFi-enabled devices and industrial transmitters and devices. Mobile operator station is such a device with WiFi connection. Through WiFi access point and using mesh infrastructure, these operator stations can access the main data server of DCS in the control room. The application running in the mobile device gives the same look and feel of the operator stations used for controlling and monitoring inside the control room. This gives lot of flexibility and mobility for the field operators to visualize the running process from anywhere in the field.

14.7.11 VoIP solution

VoIP phones are the WiFi-enabled phones. Any smart device with WiFi feature can run an application to convert the device into a VoIP phone. This becomes an economical solution to connect thousands of people in a large plant. The field to control room communication can be achieved successfully using these phones. In addition, integrating with an EPBAX system, so that each phone has an extension number and they can communicate to each other seamlessly. All this can be achieved without any additional infrastructure investment.

14.7.12 Video solution

Live video streaming is becoming an important requirement for process plants. Starting from flare monitoring to critical process area monitoring to perimeter monitoring for security purpose, live video streaming is essential. Wireless communication can also be deployed for this purpose. WiFi-enabled cameras can directly connect to the wireless access point infrastructure and dump the data in the main backhaul. IP-based cameras can be accessed from anywhere in the plant. The main advantage of this solution is the scalability. It eliminates the need to plan cable routing whenever a camera or other accessories need to be added. A WiFi camera can be mounted and operational in few minutes. Only thing that requires attention is the throughput calculation vis-à-vis available bandwidth. 802.11a gives a maximum bandwidth of 54 Mbps and 802.11n gives up to 600 Mbps (under certain conditions). But available bandwidth can be much less based on the RSSI, distance, and number of mesh hops present between the camera and the control room.

14.7.13 RFID-based solution

RFID is a technology that uses radio waves to transfer data from an electronic tag, called RFID tag or label, attached to an object, through a reader for the purpose of identifying and tracking the object. RFID tags are small RF tags that can be uniquely identified by a WiFi-based wireless infrastructure. RFID tags are of two types – passive and active. Passive RFID tags do not need any power supply and are ideal for short-range usage. These are very economical (even less than Rs. 10 or $0.2). Active RFID tags are little expensive and come with a battery. These are ideal for long-distance application. RFID tags can be used for unique item identification. In an assembly line-based production system, a part can be easily identified. Even position can be identified through triangulation technique. Location detection can be a very important aspect of a plant operating with hazardous material from the safety point of view. In case of an emergency or an emergency evacuation, people can be located and identified.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128009390000140

A review on mercury in coal combustion process: Content and occurrence forms in coal, transformation, sampling methods, emission and control technologies

Shilin Zhao , ... Jianhong Lu , in Progress in Energy and Combustion Science, 2019

5.1.3 Mercury-continuous emission monitoring system

The mercury continuous emission monitoring system (Hg-CEMS) can monitor the real-time concentration of different Hg ^g, including Hg^g, Hg²⁺, and total Hg^g. It is mainly composed of pretreatment/conversion unit and analysis unit, the schematic diagram of which is shown in Fig. 16. At present, the mercury analyzer can only determine the Hg⁰ concentration. Thus, Hg²⁺ in flue gas should be firstly transformed into Hg⁰, and then be determined. There are two ways to transform Hg²⁺ into Hg⁰, which are (1) wet chemical reduction method using SnCl₂ solution, and (2) high-temperature catalyst or high-temperature heating conversion method.

Fig. 16. Schematic diagram of the Hg-CEMS.

Read full article

URL:

https://www.sciencedirect.com/science/article/pii/S0360128518301308

On-line spectroscopic and spectrometric methods for the determination of metal species in industrial processes

Penelope Monkhouse , in Progress in Energy and Combustion Science, 2011

5.4 Conventional emission

For mercury, commercial continuous emission monitoring systems have been evaluated (Section 4.2.1, [287,288]) at different coal combustion utilities (PPC, CFB) over several months operating time and results compared to values obtained using reference techniques, here the Ontario-Hydro (OH)-Method [205] and Method 30A, a conceptual protocol proposed by EPA as a reference for CEM systems [204]. Procedures specified by US Federal Regulations were followed. All detectors were based on atomic fluorescence, half of them using gold traps to capture elemental mercury to form an amalgam. Aspects considered included linearity, response time, day-to-day stability and efficiency of the Hg speciation. Detailed relative accuracy (RA) and bias tests were performed. Two of the CEM systems gave RA values less than 20%, but the other field tests were also considered acceptable because the deviations from the OH-method values were less than 1.0 μg/cm³ (dry basis).

The multimetal monitoring system based on X-ray fluorescence (XRF) [289, Section 4.2.1] has been evaluated in detail [290]. Field measurements were made at a diesel-fired combustor for incinerating decommissioned munition. The furnace end temperature was 538 °C, the stack temperature 260 °C and the stack gas velocity 14 m/s. The system was tested for its ability to measure Sb, Se, Ag, Tl, As, Hg, Ni and Zn in the feed stream and Pb in the stack gas. Measurements in the stack gas were also attempted for the other metals, though here spiking was needed to ensure sufficient concentrations. All results were correlated with data obtained using the EPA M29 method. For Ni, Pb, Sb and Zn, r² -values were better than 0.95, but for Hg, r² was only 0.39. Possibly, chemical reactions or deposition had occurred in the sampling lines.

Systems for on-line analysis of sampled process gas by atomic flame emission also exist and these were described in the previous review [218,p.361].

Read full article

URL:

https://www.sciencedirect.com/science/article/pii/S0360128510000407

Emissions and exposure assessments of SOX, NOX, PM10/2.5 and trace metals from oil industries: A review study (2000–2018)

Patrick Amoatey , ... Abdullah Al-Mamun , in Process Safety and Environmental Protection, 2019

6 Monitoring and estimation of the pollutants

Determining the extent of human and ecosystem damages from oil industries depends on the rigorous methods of monitoring and assessing of the ambient pollutant emissions (Kampa and Castanas, 2008; MacIntosh et al., 2010; Morra et al., 2009). The traditional method of measuring the pollutants has been limited to the application of samplers (Callen et al., 2014; Chen et al., 2017a; Lang et al., 2017; Misawa et al., 2017; Sosa et al., 2017). This method has several limitations, such as (1) it cannot measure emissions continuously, (2) it has low spatial coverage, (3) it cannot perform long-term measurements, and (4) it is expensive and labor intensive approach (Melymuk et al., 2011; Mittal et al., 2013; Ramos et al., 2018; Schindler et al., 2016; Yusa et al., 2009). Due to these limitations, AQMS has offered an opportunity to fill the gaps as it can simultaneously measure several types of air pollutants with high temporal and spatial coverage. Such specifications, making AQMS as a preferred instrument in measuring several pollutants from oil industries (Khan et al., 2017; Wu et al., 2015; Xiao et al., 2018: Zhang et al., 2013). However, measurement of ambient pollutants from oil industries with AQMS may not be the best approach due to infiltration and impact by similar pollutants from other sources (transportation, traffic, constructions) (Brand et al., 2016; Chen et al., 2018b; González and Rodríguez, 2013; Pedroso et al., 2016 ). Additionally, continuous emission monitoring system (CEMS) is a very important instrument that determines real-time emission rates or flow rates of different pollutants, especially SO ₂, NOx, and PM from stationary sources.

CEMS data of a typical oil refinery showed that the default PM₁₀ and PM_2.5 emission values for the production of 50,000 barrels of crude oil were 208 tons yr⁻¹ and 189.5 tons yr⁻¹, respectively (EPA, 2011). According to CEMS measurements, for a continuous refinery combustion operation, 3.5 tons yr⁻¹ of SO₂ and 940 tons yr⁻¹ of NOx were emitted to the ambient environment for 150 × 10⁶ British thermal units (Btu) per hour burning of refinery fuel gas (EPA, 2011). The mean emission factor for Cd, Hg, Ni and Pb in pounds per million British thermal units (lb MMBtu⁻¹) were 2.38 × 10^-3, 3.23 × 10^-4, 5.59 × 10^-3 and 2.42 × 10^-3, respectively (EPA, 2011). The emissions of these pollutants in cities, regional and global levels can be estimated according to the data shown in Table 2. Based on the emission factors, the contribution of Ni compared to the other elements is predominant (Harari et al., 2016). Additionally, psychological problems of the exposed population could be due to spillage of petroleum products (Ramirez et al., 2017).

In recent years, several AQDMs have been used as efficient and cost-effective means of determining temporal and spatial pollutant emissions from different industries. AQDMs are considered as preferred models as they could simulate air pollutants from emission sources at ground level concentrations (Amoatey et al., 2018a). The common AQDMs are Industrial Source Complex Short Term (ISCST) (Abdul‐Wahab et al., 2008), California Puff model (CALPUFF) (Lang et al., 2017; MacIntosh et al., 2010; Rood, 2014), American Meteorological Society/US Environmental Protection Agency Regulatory Model (AERMOD) (Chen and Carter, 2017c; Michanowicz et al., 2016) and Hybrid-Single Particle Lagrangian Integrated Model Trajectory (HYSPLIT) (Dodla et al., 2017). The basic data required in estimating pollutants with these models include emission rates, stack data (height, diameter and temperature), local meteorological information (ambient temperature, wind speed and direction, sensible heat flux, and convectional layer), and land use characteristic data (elevation, albedo, etc.) (Abu-Eishah et al., 2014; Amoatey et al., 2018a, 2017). The measurement instruments such as passive samplers, AQMS, and CEMS could be modified to measure/validate pollutants from oil industry emissions compared to AQDMs.

Amoatey et al. (2018a) reported that employing AQDMs in assessing ambient concentrations of SO₂ and NO₂ emissions from an oil refinery within a residential location could be very beneficial in assessing future health risk assessment (HRA) of the residents due to the reliability of the model. Comprehensive data about several ambient measurement procedures and their associated measured pollutants from the selected oil industries are shown in Table 3. The table contains wide variety of pollutants, which are classified by US EPA (2018b) and European Commission (2016) as human carcinogenic. The major results from the table are: 1) NO emission as a primary ambient pollutant, is less considered in the research studies. According to Wu et al. (2019), NO is the main component of NOx in combustion flue gas (˜95 Vol.% NOx), which plays a critical role in the atmospheric environment research, and 2) many of the high capacity refineries that can emit these pollutants are located in countries such as Saudi Arabia, Iran, Russia, Nigeria, China, and Venezuela, whereas most of the pollutants are not properly monitored and regulated; thereby, they may be exceeded most of the international and global standards (Hadidi et al., 2016).

Table 3. Ambient pollutants emission from selected oil industries in different countries.

Study Location	Industry Unit	HRA	Instrument/ Model Used	Pollutants Emitted (GLCa)	References
Fahaheel, Kuwait	Ahmadi	No	AERMOD	SO2 (2.5 × 10−4) ppm NO (1.64 × 10−6) ppm	AL-Haddad et al. (2012)
Umm Alhyman, Kuwait	Shuaiba	No	AERMOD	SO2 (1.52 × 10−4) ppm NO (4.37 × 10−4) ppm	AL-Haddad et al. (2012)
Al Sahil, UAE	ADGASb	No	AERMOD	24 h SO2 (124.9 μg m−3)	Abu-Eishah et al. (2014)
Sailing Club, UAE	ADGAS	No	AERMOD	24 h SO2 (111.06 μg m−3)	Abu-Eishah et al. (2014)
Al Jimi, UAE	ADGAS	No	AERMOD	24 h SO2 (194.92 μg m−3)	Abu-Eishah et al. (2014)
Montreal, Canada	Nil	Yes	Ogawa samplers	24 h NO2 (6.33 ppb) 24 h PM2.5 (9.6 μg m−3)	Smargiassi et al. (2014)
Quebec Provinces	Nil	Yes	Monitoring stations	Annual PM2.5 (177.37 ton) Annual SO2 (2327.23 ton) Annual NO2 (773.0 ton)	Brand et al. (2016)
British Columbia	Nil	Yes	Monitoring stations	Annual SO2 (577.79 ton) Annual NO2 (164.34 ton)	Brand et al. (2016)
Taiwan	Petro-AQS	No	AQMS	24 h SO2 (24 ppb) 24 h SO2 (4.5 ppb)	Chen et al. (2018b)
Dacheng, Taiwan	No.6 Naptha cracking complex	Yes	Urinary Biomarker application and ICP-MS	As (103.76 μg g−1) Ni (6.88 μg g−1) Pb (1.54 μg g−1) Hg (1.38 μg g−1)	Chen et al. (2018a)
Zhutang, Taiwan	No.6 Naptha cracking complex	Yes	Urinary Biomarker application and ICP-MS	Cr (1.94 μg g−1) Cu (9.97 μg g−1) Cd (0.8 μg g−1) V (0.49 μg g−1)	Chen et al. (2018a)
Taisi, Taiwan	No.6 Naptha cracking complex	Yes	Urinary Biomarker application and ICP-MS	Sr (140.76 μg g−1) Mn (3.52 μg g−1) Hg (1.38 μg g−1) Pb (1.67 μg g−1)	Chen et al. (2018a)
Taishi, Taiwan	No.6 Naptha cracking complex	No	192 PM10 filters, ISC3	As (7.5 ng m−3) V (2.23 ng m−3)	Chio et al. (2014)
Taishi-Yulin County, Taiwan	No.6 Naptha cracking complex	No	Monitoring stations, ICP-MS	PM2.5 (30.1 μg m−3) Mn (7.01 ng m−3)	Chuang et al. (2018)
Beijing, China	Nil	No	Vacuum 3 L-Summa Canister Sampler	1 h NO2 (46.1 ppb) 1 h NO (6.4 ppb)	Wei et al. (2014)
Tema, Ghana	Tema Oil Refinery	No	CALPUFF	24 h SO2 (88 μg m−3) 24 h NO2 (9.6 μg m−3)	Amoatey et al. (2018a)
Tema, Ghana	Tema Oil Refinery	Yes	AERMOD	24 h PM2.5 (38.8 μg m−3) Annual PM2.5 (12.6 μg m−3)	Amoatey et al. (2017)
Aliaga Region, Turkey	Nil	Yes	Passive sampler	24 h SO2 (21.1 μg m−3) 24 h NO2 (17.8 μg m−3)	Civan et al. (2015)
Algeciras and La Linea, Spain	San Roque Refinery complex	No	Cascade impactor sampler	Cr (962 μg m−3), Zn (225 μg m−3) V (638 μg m−3), Ni (3295 μg m−3) Mo (91 μg m−3), Co (94 μg m−3)	de la Campa et al. (2011)
Santa Cruz de Tenerife	Nil	No	Ultrafine Condensation and Optical Particle Counter, Monitoring station	SO2 (10 μg m−3), NO (56 μg m−3), BC (1035 ng m−3), PM10 (26 μg m−3), PM2.5 (13 μg m−3)	González and Rodríguez (2013)
Asaluyeh, Iran	South Pars Complex	No	AERMOD	1 h NO (700 μg m−3)	Jafarigol et al. (2015)
Faheel and Al-Riqa, Kuwait	Mina Abdullah Refinery	No	Mobil Air Monitoring Lab	PM10 (297 ppb), NOX (21 ppb) H2S (14 ppb), NO2 (12.2 ppb)	Khanfar (2015)
Uslan, Korea	Ulsan Petrochemical Industrial Complex, On-San Industrial Complex	No	WRF-CALPUFF	24 h PM10 (157.7 μg m−3)	Lee et al. (2014)
La Linea and Puente Maryoga,Spain	CEPSA Oil Refinery	No	PM10 Sampler	24 h PM10 (31 μg m−3)	Li et al. (2018a)
Asaluyeh, Iran	Nil	No	AERMOD	Annual NO2 (217.4 μg m−3)	Minabi (2017)
Sao Paulob, Brazil	Nil	No	Monitoring station	Hourly PM10 (785 μg m−3)	Nakazato et al. (2015)
Cubataoc, Brazil	Nil	N/A	Monitoring station	24 h NO2 (37.3 μg m−3), SO2 (11 μg m−3), PM10 (37 μg m−3)	Pedroso et al. (2016)
Altamira, Mexico	Nil	No	PM10 Sampler	24 h PM10 (92 μg m−3)	Rodriguez-Espinosa et al. (2016)
Thailand	Nil	No	AERMOD	1 h SO2 (359 μg m−3)	Thepanondh et al. (2016)
Thailand	Nil	No	CALPUFF	1 h SO2 (456 μg m−3)	Thepanondh et al. (2016)
Argentina	Nil	No	TOEM monitors	24 h PM10 (50.7 μg m−3)	Singh et al. (2015)
Chikun, Nigeria	Kaduna oil refinery and petrochemical complex	No	ISCST3	24 h PM10 (0.4 μg m−3) 24 h SO2 (82.7 μg m−3) 24 h NOX (164.1 μg m−3)	Oladimeji et al. (2015)
Eleme, Nigeria	Port Harcourt Refining Company (I and II)	No	ISCST3	24 h PM10 (0.45 μg m−3) 24 h SO2 (69.1 μg m−3) 24 h NOX (1855.25 μg m−3)	Oladimeji et al. (2015)
Warri, Nigeria	Warri refining and petrochemical company	No	ISCST3	24 h PM10 (1.7 μg m−3) 24 h SO2 (444.4 μg m−3) 24 h NOX (2145.8 μg m−3)	Oladimeji et al. (2015)

^aGLC is the major concern for practical environmental and health assessment studies.

^bADGAS: Abu Dhabi Gas Liquefaction Company.

^cReceptor: Agricultural field.

^dReceptor: Forest.

Read full article

URL:

https://www.sciencedirect.com/science/article/pii/S0957582018313557

ricethimpubstur.blogspot.com

Source: https://www.sciencedirect.com/topics/engineering/continuous-emission-monitoring-system