Dorset - Biological Air Washer for strawless pig houses

* The DLG test framework provides the following options in its evaluation schemes: 
■■■ or better = meets, exceeds or clearly exceeds the specified DLG standard,
■■= meets the legal requirements for marketability,
■ = failed

**Evaluation range: Meets the requirement  (✔) /  Does not meet the requirement (X)

[1]    Three total dust measurements were carried out in 2005, the results of which met the DLG requirements. The measurements shown here were obtained in the 2020 follow-up test (winter and summer).
[2]    The minimum separation rate for dust is the lowest measurement recorded in the test period. The minimum ammonia separation efficiency is the averaged separation efficiency minus the standard deviation.
[3]    Experience has shown that the scrubbing process can lead to the formation of droplets of 2.5 µm to 10 µm in size. This in turn leads to an increase of PM10 particles in the cascade impactor. The PM2.5 particle fraction is not affected as much by this effect and therefore the separation efficiency for this fraction is calculated to be higher than for the PM10 fraction.
[4]    As the concentration of pollutants in the waste air was low in the summer, it was not possible to obtain valid measurements under summer conditions.
[5]    The clean air measurements in the summer were corrected as per DLG test framework in that ammonia concentrations of less than 1.0 ppm were raised to 1.0 ppm.
[6]    One winter measurement on N removal was not plausible. This was probably attributed to inaccurate air flow measurements due to low flow rates or inaccurate water level measurements. 
[7]    All annual consumption data refer to 365-day system uptime. This approach allows the system to be compared with other systems. 
The actual consumption may be lower due to service and downtime periods.
[8]    Blowdown was always carried out when conductivity reached its maximum of 20 mS/cm.
[9]    These consumption measurements refer to the substance nitrapyrin (17.7 %) of which only relatively small quantities were consumed by the reference system. These rates may be higher in real-life operations.

Description and technical data

The “Biological Air Washer” is a single-stage system with an upstream spray nozzle and a downstream droplet separator for cleaning waste air emissions from strawless pig houses with overhead ventilation. This system is operated on the pressure principle. 

The waste air passes from the animal area into the waste air collecting chamber and is fed by synchronised fans into the scrubbing area after it is sprayed with water for an initial removal of coarse dust particles and humidification. After the waste air enters the scrubber from the bottom it makes a 180° upward turn and is pressed vertically through the 90 cm packing that is arranged at the top. The biological processes that take place inside this packing oxidise ammonia and odours and remove these from the air flow. The packing is sprinkled with water by a system of nozzles that are arranged above the packing. The water pH is between 6.8 and 7.1 and is controlled by adding acid and a nitrification inhibitor. The water is collected in a sump under the packing. This process/circulating water is pumped back to those nozzles that sprinkle the filter bed.
The filter bed operates on the principle of biological oxidation by intensive interaction of the pollutants with the circulating water on the large 240 m²/m³ specific surface of the packing. The surface areas in the packing are clustered by microorganisms that metabolise the pollutants and generate new biomass, forming a film of biomaterial. After passing this biological stage, the cleaned air leaves the scrubber, passes through a droplet separator and exits to outdoors. The droplet separator retains the aerosols, thereby reducing the amount of water that would otherwise be lost.

The fans are synchronised, which means all feed identical waste air volumes. The scrubbing system is designed for processing a maximum filter surface load of 1,980 m³/(m² · h). The packing must constantly be sprinkled with process water. This is necessary to ensure a high biological activity and smooth system processes. The sprinkling density for the filter bed must be  at least 0.71 m³/(m² · h) in addition to another 0.06 m³/(m² · h) that is generated by the nozzle at the entrance of the scrubber.

The process water pH is between 6.8 and 7.1; its mean half-hour level must be logged to the electronic logbook (ELB). When pH exceeds the permitted maximum, acid is metered into the process water to reduce pH. When pH drops below minimum, a nitrification inhibitor is added to prevent pH from dropping further. As more ammonia enters the systems with the waste air and nitrite and nitrate formation is inhibited, pH increases again. When pH reaches 6.5, the system automatically starts metering the nitrification inhibitor into the water. This process is stopped when pH is back at 6.8. As ammonia is separated and nitrification is taking place the concentration of these particles in the process water increases. The level of pollutants that have accumulated in the process water is measured by a sensor that determines the inductive conductivity of the liquid. This information is logged as mean half-hour values (mS/cm) to the electronic logbook.

For an effective removal of nitrogen in a single-stage biowasher, blowdown is carried out when conduc­ti­vity reaches 20 mS/cm. This is an automatic routine during which freshwater is topped up as well. Topping up dilutes the process water and reduces its conductivity.

Since scrubbing also leads to increased evaporation, it is necessary to enter the freshwater and blowdown figures to the electronic logbook. The water level 
is measured electronically by a level sensor. This sensor also helps protecting the circulating pumps from running dry.

After the biological air scrubber has been installed, it takes six to eight weeks before the system achieves its full effectiveness in reducing odour particles. This does not apply to ammonia separation and N removal which are pH-controlled processes that set in immediately. The same applies to dust separation. 

This means that it is necessary to start controlling the pH level of the process water on the very first day of operation. The same applies when resuming operation after downing the system in the pig finishing or piglet production cycle or for cleaning. To avoid longer start-up phases after the system was downed (not after cleaning), it is necessary to con­tinue sprinkling the packing during any downtime. 

The first test was carried out in a reference plant in the Netherlands (2005 summer measurements). Some of these measurements were approved to District Cloppenburg approval procedures in the same reference operation (winter measurements). The test results were discussed in the DLG Test Report 5702 which is now superseded by this Test Report.

After a number of innovations were introduced to the system (conductivity-based blowdown, process water pH control, etc.), the company applied for repeating the measurements in 2019. These were carried out in winter and summer 2020. The measuring periods were two four-week periods that were agreed on by the DLG Test Commission. The reference system was installed on a pig farm in Lower Saxony. This farm was a 1,100-head finisher operation. The pigs were finished in a continuous operation in strawless houses (entering / exiting in groups). 

As this type of system was not yet on the market at the time of testing, it was not possible to survey potential owners of this type of scrubber in parallel to the tests.  The scrubber is approved for operating on the pressure principle. The measurements were taken from March to April 2020 (winter measurements) and from June to July 2020 (summer measurements). 

The evaluation of the scrubber covered the following parameters:

Dust

The samplings followed the VDI Guideline 2066, Sheet 1 and DIN EN 13284-1. For this purpose, an isokinetic sampling system to Paul Gothe with a 50 mm diameter plane filter head device was installed. A 45 mm diameter glass fibre circular filter served as separating medium.

The fine dust measurements (PM10 and PM2.5) were carried out to VDI Guideline 2066, Sheet 10 and DIN EN ISO 23210. The instrument used was a cascade impactor Johnas II to Paul Gothe with three 50 mm diameter plane filters. The separating medium was again a glass-fibre circular filter, this time of 50 mm diameter. The performance was assessed by measuring the dust load gravimetrically.

The DLG test framework stipulates that the separation efficiency must at least be 70 %. This applies to total dust and fine dust (PM10 fraction). The results of the PM2.5 measurements are merely reported here but not assessed. The minimum separation rate is the smallest rate obtained in all measurements on the sampling days.

Ammonia

The ammonia measurements taken in the waste air and clean air areas of the system were carried out continuously over the entire sampling period by applying FTIR spectroscopy in accordance with KTBL publication 401 and DIN EN 15483, using a measuring cell. To prevent condensation in the sampling PTFE lines, the lines in the clean air system were heated down their full lengths. The plotted results reflect actual measurements. Whenever these were less than 1.0 ppm, the value was raised to 1.0 ppm. This is justified by the measurement uncertainty of the instrument. It is not possible to produce a precise measurement below this limit.

For proving compliance with the Animal Welfare guidelines (TierSchNutztV) (max. 20 ppm NH3 in the animal area), the ammonia concentration was coninuously measured at animal level on the sampling days.

For obtaining DLG approval to the DLG test framework, the system must be able to separate consis­tently more than 70 % from the exhaust air. This 
rate is verified by averaging all measurements and deducting the standard deviation figure from this averaged result. 

Aerosol emissions

By sprinkling a scrubber packing with water, the nitrogen-laden aerosols are expelled as ammonium aerosols from the packing and are carried away by the exhaust air flow. This way, nitrogen (N) that was initially separated now returns unintentionally into the ambient air. 

The aerosol emissions into the ambient air during the sampling period were measured by a plane filter, with one measurement carried out in the winter period and another in the summer period. For this purpose, two sampling instruments were installed, one of which had a particulate inlet filter that retained the aerosols. The samples were taken in line with VDI Guideline 3496-1 (on measuring gaseous ­emissions).

According to the DLG test framework, the aerosol output must not exceed 0.50 mg nitrogen per standard cubic metre. 

Nitrogen balance and nitrogen removal

The measurements of ammonia removal from the waste air were verified with the help of an N balancing method by considering the ammonia load in 
the waste air and the cleaned air as well as the inorganic nitrogen compounds in the wastewater. The N balancing period in the summer was 28 days. In the winter it was not possible to obtain a plausible N balance, because the flow rates were low and the measurements not accurate enough.

In biological washer systems, the process water must be sampled for ammonium (NH₄-N), nitrite (NO₂-N) and nitrate (NO₃-N).
The method for determining the quantity of actually removed N is to ascertain the ratio of inorganic N mass that is removed from the system and of the N load that enters the system with the waste air. 

Balancing the nitrogen flows within the plant is important for two reasons: 

• because all relevant nitrogen compounds and their whereabouts are documented, 
• because the nitrogen content of the blowdown wastewater is determined and its fertiliser levels are quantified. 

The DLG test framework stipulates a nitrogen balance and N removal rate of at least 70 % during the winter and summer measurements.
The test framework stipulates a nitrogen recovery rate (N balance) of 80 % minimum and 120 % maximum.

Odour

The odour concentration was determined with the help of dynamic olfactometry to DIN EN 13725, using the yes/no method by diluting the odour down to a threshold where it is no longer perceived. 

The sampling of waste air and cleaned air was carried out with the help of a negative pressure sampler (CSD-30). The sample was analysed by Ecoma GmbH by means of a T08 olfactometer. 

The olfactory acuity of the test engineers was tested using n-butanol as standard odorant at each testing. In each test period, the odour removal efficiency was determined by taking eight odour samples. 

The DLG test framework stipulates that the odour concentration must not exceed 300 OU/m³ in any of the measurements and that no typical raw gas odour (pig) may be perceptible in the sample.

Consumption values, ambient conditions and system load

The consumption of freshwater, blowdown waste­water and electrical energy was measured by recording the readings on the electricity meters – one on the cleaning system and one on the ventilation system. 

The consumption of acid and nitrification inhibitors was determined with the help of a weighing system (sensor or load cell or scales).  The ambient conditions (temperature and relative indoor/outdoor humidity) were also recorded during the measurements. On the days when dust and odour concentrations were measured, the following parameters were also recorded: 

• Animal weights (estimated weights) and numbers of animals
• Freshwater and power consumption (meter readings)
• Absolute air flow (calibrated measuring fans and separate fan curve)
• Pressure loss across the scrubber and pressure loss across shed and scrubber
• Process water pH and conductivity 

Furthermore, the measurements that had been logged in the logbook by the manufacturer were verified for plausibility.

Operational reliability and durability

Operational reliability and durability were also assessed in this test. Any malfunctions of the system as a whole or of technical components during the test period were documented.

Operating instructions, handling and operation, workload and maintenance requirements

The detailed description of the system and its way of operation as well as the clear descriptions of regular service and maintenance jobs were assessed from a user’s point of view. Tests on handling and operation as well as workload assessed whether it is necessary for the manufacturer to provide initial training when commissioning the plant and the time spent on routine inspections and on daily, weekly and monthly routines as well as on troubleshooting.

In terms of service and maintenance requirements, the testers review the service intervals and checklists.

Documentation

The following parameters must be recorded and saved to the electronic logbook as half-hourly means or half-hourly readings:

• Pressure loss across the scrubber [Pa]
• Air flow rate [m³/h]
• Pump runtime (circulating water, blowdown) [h]
• Sprinkling rate [m³/h]
• Accumulated freshwater consumption [m³]
• Accumulated blowdown volume [m³]
• Waste and clean air temperatures [°C]
• Process water pH [-] and electrical conductivity [mS/cm] as half-hourly means
• Accumulated power consumption [kWh]
• Accumulated consumption of nitrification inhibitors [kg]

In addition, spray patterns, service and repair times, plus pH and conductivity sensor calibrations must be recorded. The same applies to the consumption of chemical additives (e.g., acid, defoamer). These data reflect whether the air scrubber is  operated to instructions; they were verified on the reference system.

Environmental safety

The test module on environmental safety comprised the assessment of all inputs for operating the system (e.g., acids, alkalis, nitrification inhibitors), the use of waste products such as the blowdown wastewater and the replacement and disposal of individual parts of the plant. It was also examined who would be in charge of the various aspects.

Safety aspects

To assess the safety of the system, the plant was inspected for compliance with the current fire and plant safety requirements.

Dust

The particle separating technology (trickle bed filter) and the upstream nozzle at the entrance to the scrubber met the requirements as laid down by the DLG test framework.

During the winter period, two total dust and two fine dust measurements were conducted (PM₁₀ and PM_2.5). Two total dust measurements and two fine dust measurements were also conducted in the summer, yet the fine dust measurements under summer conditions could not be used for the assessment. This was attributed to the fact that the flow rates were very high in this test period – to the effect that the waste air levels in the air flow were too low for obtaining valid separation efficiency rates.

The smallest separation rate of total dust was 80.8 % in the winter and 71.2 % in the summer. The smallest separation rate of fine dust particles (PM₁₀) was 83.3 % in the winter.

Experience has shown that the scrubbing process can lead to the formation of droplets that range between 2.5 µm and 10 µm in size. These lead to an increase of PM₁₀ particles in the cascade impactor. The PM_2.5 particle fraction is not affected as much by this effect. As a result, the separation efficiency for this fraction is calculated to be higher than for the PM₁₀ fraction. 

Ammonia

For ammonia separation in the Biological Air Washer to meet the minimum requirements the process water must be drained automatically when conductivity reaches the maximum of 20 mS/cm and water pH is maintained between 6.8 and 7.1.

In the winter measurements, 1619 half-hourly mean results were available as paired values and 1816 in the summer measurements. Regular measurements of ammonia concentrations at animal level produced more or less consistent results Only few temporary exceedances by more than 20 ppm were observed in the winter measurements, which were down to animal activity. The mean ammonia concentration was 12.1 ppm. There were no exceedances of those rates specified by the animal welfare TierSchNutztV ordinance in the summer.

In the winter measurements the minimum separation efficiency was 83.5 %, in the summer 83.9 %. 

Odour

Since the test was a follow-up test of an existing yet updated system, the DLG test committee unanimously decided on a reduced test programme which consisted of two measuring days in the winter and two in the summer.

With raw gas concentrations of 1,390 OU/m³ in the waste air in the winter and up to 1,642 OU/m³ in the ­summer, the testers measured less than 300 OU/m³ in the cleaned air on all four testing days. They never perceived a raw gas odour. 

In a third measurement in the winter, the waste air measurement produced a concentration of 3000 OU/m³ after slurry in the pit below the main aisle was agitated. On that day, the concentration of 300 OU/m³ in the cleaned air was exceeded and raw gas odour was perceived in the cleaned air. Any pig producing operation must achieve a concentration of ≤ 300 OU/m³ and no perception of raw gas odour. Agitating slurry in the pit or emptying the pit leads to a temporary increase in odour emissions which the scrubber cannot fully remove. For these emissions to be fully filtered and removed, the air would have to dwell much longer inside the scrubber, for which this would have to be larger. This in turn would not be economical since this is only a temporary occurrence. 

On the whole, the criteria in the DLG test framework were complied with (≤ 300 OU/m³ and no raw gas odour). In all samplings either all or most of the testers didn’t perceive any raw gas odour in the cleaned air.

The results of the odour samplings are listed in table 4.

Consumption rates, ambient conditions and system load

Table 1 also lists the results of the consumption measurements per measuring period (winter/summer). The figures listed here are scaled to reflect annual consumption rates (365 days) to make them comparable with data from other brands. Since some of these rates vary significantly (winter/summer differences), only the average consumption (averaged winter and summer consumptions) is discussed in the following evaluation. These consumption rates are converted to consumption per livestock place and year (LP · a). This calculation is based on the number of 1,100 finishers for which the reference plant is approved.

Water consumption

The water consumption figure is based on the frequency at which blowdown is carried out and on evaporation levels. The higher the blowdown frequency and the evaporation rates, the more freshwater needs to be added to the system to maintain the required water level in the system. The blowdown frequency is controlled by the nitrogen input through the waste air flow on the one hand and process water conductivity on the other. When the latter reaches 20 mS/cm, blowdown is initiated. During the test period, an average of 1.5 m³ of blowdown wastewater was drained from the sump every day. The average annual blowdown per livestock place is 0.543 m³.

The total consumption of freshwater was 3.8 m³/d and 1.42 m³/(LP·a). The freshwater was added directly to 
the sump.

The blowdown and freshwater volumes compare with typical volume rates in other facilities.

Power consumption

The circulation pump is operated permanently and is the chief electrical consumer in the Biological Air Washer. The greatest consumers in the pig house are the fans. The reference plant had three pressure-resistant exhaust fans. Each fan was connected to a frequency converter which controlled the fan speed relative to the current waste air flow.

The pressure loss across the cleaning system never exceeded 10 Pa in the packing and 5 Pa in the droplet separator; hence a combined 15 Pa across the scrubber. The maximum pressure loss recorded for the shed and the scrubber was 63 Pa. This means that the fans for the shed and scrubber must have the capacity to handle a pressure loss of at least 70 Pa. In the tested shed the fans were slightly undersized.

Their annual mean consumption was 44.1 kWh/day; yet consumption in the summer was much higher than this. The above figures translate into an electric consumption of 16.8 kWh per livestock place and year.

The annual average power consumption by the scrubber was 68.5 kWh/day or 25.3 kWh/(LP · a). The reason why power consumption was slightly higher in the summer was because the sprinkling density was lower in the winter. The specific power consumption may drop as the size of a plant increases.

Further consumption rates

For the system to operate reliably and at the efficiencies depicted, pH must range between 6.8 and 7.1 and blowdown is carried out when conductivity reaches 20 mS/cm. Therefore, it is necessary to have an automatic acid dispenser in place and a conductivity sensor in each stage of the cleaning process. The pH level in the reference plant was reduced by adding 96 % sulphuric acid.

The measurements produced an average annual acid consumption of 5.8 kg/d which translates into 
2.21 kg/(LP · a).

To prevent pH from dropping below 6.8, a nitrification inhibitor must be added. The inhibitor used in the reference plant was N-Lock with the active substance nitrapyrin (17.7 % mass fraction). 

The test showed that N-Lock consumption in the winter was zero. The average annual consumption of nitrapyrin was 16.9 g/d. This translates into 6.7 g per livestock place and year. This is relatively little compared with other systems. Using the information on the delivery dockets from the testing period, the test engineers calculated an annual consumption of nitrapyrin of 16.4 g/(LP·a), which seems a more realistic rate.

N-Lock is not the only nitrification inhibitor that can be used in scrubbers. Yet, as no other agents were assessed in this test, we cannot comment on those rates.

Operational reliability and durability

During the test period, the system technology didn’t experience any malfunctions that would be worth mentioning. The same applied to the scrubber as a whole which didn’t experience any damage or wear symptoms worth mentioning during the test.

As far as it was possible to determine during the test period, the individual components of the plant seemed to be properly protected against corrosion. The system is a complete system and nearly exclusively made from plastic. The only exception is the fire protection plate girder which is made from stainless steel.

This comment on durability applies only to the testing period. Since the plant was operated as a pilot project, it was not possible to carry out a survey among operators of similar plants.

Manual, handling and operation, workload, service and maintenance requirements

The manual is comprehensive and explains the system in great detail. It also details on daily, weekly and annual service intervals. Photos of the various elements of the plant visualise the instructions. 

Operators must receive a training by the manufacturer and understand the instructions.

After the scrubber has been taken into operation and operated for a sufficiently long time, it can be con­sidered as being easy to operate, because in standard mode the scrubber runs fully automatically. The only servicing required by operators is a daily check of the operating data and a weekly inspection of the entire scrubber including the nozzles. 

Error codes issued on the control unit are explained in the manual and instructions are given on how to inspect the system element in question. It is advised to sign a maintenance agreement with the manufacturer to save time and avoid handling issues.

When the spray pattern becomes irregular, the nozzles need cleaning or replacing. Therefore, they have to be inspected on a weekly basis. These inspections must be logged to the logbook.

Documentation

All data that are required for the safe operation of the system are regularly logged to the electronic logbook as half-hourly mean values (pH and conductivity) or half-hourly values. Logging is automatic and the data are stored for 5 years. They can be exported by the operator, the manufacturer and authorities – either remotely or to a USB drive and then viewed in a common table format. The type of data recorded is shown in table 7. 

If the ventilation system in the shed is sourced from a different manufacturer than the air scrubber, the manufacturer of the scrubber will compute the ventilation data into a characteristic curve and integrate it in the scrubber control system. The maximum fan capacity is calibrated to 100 %. It is however not adapted to another range of performance. Since the test framework and the TA Guideline on air cleaning systems (Section 5.4.7.1) require the air flow rate to be indicated in absolute numbers and in m³/h format, a characteristic curve of the entire ventilation system (in the shed and in the scrubber) must be computed and logged to the electronic logbook before the plant is taken into operation. The characteristic curve should mark out at least five different sampling points between 0 % and 100 % air flow.

Testing agency
DLG TestService GmbH, Standort Groß-Umstadt 
The tests are conducted on behalf of DLG e.V.

Laboratory and emissions measurements
SGS Institut Fresenius GmbH, 
Im Paesch 1a, 54340 Longuich

BUB Braunschweiger Umwelt-Biotechnologie GmbH,
Hamburger Straße 273 a, 38114 Braunschweig

DLG test framework
DLG-APPROVED Full Test “Exhaust air cleaning systems for livestock houses”, Date of issue March 2016

Department
Agriculture

Project Manager
Dr. Ulrich Rubenschuh

Test engineer(s)
Dipl.-Ing. (FH) Tommy Pfeifer*

Test commission
Friedrich Arends, Chamber of Agriculture of  Lower Saxony 
Christian Dohrmann, farmer
Doris Düsing, LK Cloppenburg 
Bernhard Feller, Chamber of Agriculture North Rhine-Westphalia
Ewald Grimm, KTBL Darmstadt
Dr Jochen Hahne, TI Braunschweig
Andreas Schlichting, TÜV Nord Hamburg
Thomas Üffing, farmer

*    Author