|
Whole Farm Performance of Centrifuge Extraction of Phosphorus from Dairy Manure
C. Alan Rotz1,*, Michael R. Reiner1, Sarah K. Fishel1, Clinton D. Church1
Published in Applied Engineering in Agriculture 38(2): 321-330 (doi: 10.13031/aea.14863). 2022 American Society of Agricultural and Biological Engineers.
1 USDA, ARS, University Park, Pennsylvania, USA.
* Correspondence: al.rotz@usda.gov
The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License https://creative commons.org/licenses/by-nc-nd/4.0/
Submitted for review on 20 September 2021 as manuscript number NRES 14863; approved for publication as a Research Article by the Natural Resources & Environmental Systems Community of ASABE on 2 February 2022.
Mention of company or trade names is for description only and does not imply endorsement by the USDA. The USDA is an equal opportunity provider and employer.
Highlights
Abstract. As the size of dairy farms has increased, feeds produced on the farm as well as those purchased from off-farm sources can be transported long distances to feed the herd. Transporting the manure back to the cropland used to produce the feed can be difficult and uneconomical. Technology such as a centrifuge can be used to extract nutrients into a more concentrated form for more efficient transport. A dairy farm with 2000 cows and 1400 ha of land in Pennsylvania was simulated with the Integrated Farm System Model to evaluate the feasibility of extracting phosphorus (P) to reduce transport requirements on farm or to produce a concentrated P product for off-farm use. On this farm where manure must be transported to distant cropland to obtain uniform distribution, P extraction with a centrifuge provided a better ratio of nitrogen and P contents for use on nearby cropland and reduced transport costs for nutrients applied to more distant cropland. The centrifuge was found to be more practical and economical when used with manure scraped from the barn floor than with flushed manure because much less material was handled. Moving less material through the centrifuge both improved extraction efficiency and reduced electricity consumption, providing more economical P extraction. To avoid long-term accumulation of soil P on the farm with less land (2000 cows and 1100 ha) where concentrate feed (27% of total feed) was imported, centrifuge extraction provided a material with a high P concentration exported from the farm for other uses. Extracting the P for off-farm use cost about $2.51/kg P, which was greater than the price of phosphate fertilizer, but the extract also included other nutrients and micronutrients of value to crops. A centrifuge provides a useful tool for extracting and concentrating manure P, but the economic benefit to the producer depends upon the value of the full array of nutrient contents in the product, other manure handling practices, and the end use of the extracted material. Reducing the risk of eutrophication of surface waters provides additional benefit to society.
While animal manure contains nutrients [primarily nitrogen (N), phosphorus (P), and potassium (K)] and organic material that are beneficial for crop production, the concentrations of those nutrients are often too low to allow economically viable transportation of bulk manures over long distances. In addition, dairy manure is typically in liquid or slurry forms that must be transported in tanks, so hauling for land application is inconvenient. Therefore, dairy manure tends to be applied to soils near where it is generated. Since P is conserved during manure handling compared to N and the N:P ratio in dairy manure is low, (usually between 2:1 and 10:1; Pagliari and Laboski, 2012, 2013), P concentrations in soils where dairy manure is applied tends to exceed crop demands (Sharpley et al., 1994). Due to the implication that P runoff from agricultural operations plays an important part in eutrophication of streams and other water bodies, farmers are experiencing increasing pressure and regulation to not apply animal manures to fields that are already overloaded with P (Kleinman et al., 2012).
A possible solution to the P overloading that happens when unneeded P is added to agricultural soils is to remove some of the P from manure before it is applied. As part of testing the MAnure PHosphorus EXtraction (MAPHEX) System, Church et al. (2016, 2017, 2018) showed that by treating liquid dairy manure with a screw press followed by a decanter centrifuge, 38% to 60% of the P could be removed from the manures of a wide variety of farms. A benefit to this approach is that the P removed, along with some organic material and other nutrients, is concentrated into a stackable solid (~72% moisture) that can be more economically transported to distant fields where P may be in deficit. The remaining liquid and course solids, containing greater than 90% of the manure N, can be beneficially used nearer the source without loading those soils with P.
While some farms are currently using the combined screw press and decanter centrifuge approach to recover bedding material and remove P from manures, the whole-farm implications of this approach have not been fully evaluated. A comprehensive evaluation is needed to assess the performance, environmental benefits, and economics at the farm scale. This type of assessment can be done using the Integrated Farm System Model (IFSM; USDA, 2021). The IFSM has been used to assess other manure handling strategies along with many options for crop, animal and feeding management (Veltman et al., 2018, 2021; Rotz et al., 2021a).
Our objective was to evaluate the whole-farm performance, environmental effects, and potential economic benefit of extracting P from dairy manure using a decanter centrifuge. Addition of the centrifuge was compared to a base system that used a rotary screen and screw presses to remove solids in a liquid manure flush system. Centrifuge extraction was also evaluated with manure handling converted to a scrape system which collected a thicker slurry manure. Evaluations included the base dairy system with an overall P balance for the farm, and an alternative system with a smaller land base where more feed was imported. With more manure P available on the farm than could be used by the feed crops, extraction provided a material for exporting excess P from the farm.
Materials and Methods
The P extraction process was assessed using three manure handling strategies on the same representative farm. Alternative strategies included replacing a flush manure collection system with scraping to reduce the volume of material handled by the separation equipment and removal of corn land used to produce grain which required the purchase of more feed creating excess P on the farm. Each of the farm scenarios was simulated with the IFSM both without and with the use of a centrifuge. Farms were simulated over 25 years of historical State College, Pennsylvania, weather (1991 to 2015). Various environmental impacts and manure handling costs were compared to study the potential benefits and economics of using the P extraction process. Environmental impacts included N losses, P runoff losses, fossil energy use, and greenhouse gas emissions. Manure handling costs included the amortized annual ownership cost of equipment and facilities, and annual costs for repair and maintenance, fuel, electricity, and labor.
Farm Description
The simulated farm represented an actual farm in Pennsylvania. The terrain of the region consisted of ridges with long and narrow valleys between them. Most dairy farms in this region are relatively small to avoid long travel distances for hauling harvested feed crops and returning manure nutrients to the cropland. This relatively large farm was spread over a large area with hauling distances over 7 km. Transporting manure over this distance was challenging, which created an incentive for condensing manure nutrients for easier and more economical transport.
| Table 1. Characteristics of simulated dairy farms in central Pennsylvania. | ||||
| Characteristic | Units | All-feed Farm[a] | Forage only Farm[a] | |
| Cattle | ||||
Breed | Holstein | Holstein | ||
Cows | 2000 | 2000 | ||
Heifers over a year old | 900 | 900 | ||
Heifers under a year old | 1000 | 1000 | ||
Milk production | kg/cow/ year | 11,200 | 11,200 | |
Replacement rate | % | 38 | 38 | |
Cow ration protein content | % | 17.5 | 17.5 | |
Cow ration phosphorus content | % | 0.35 | 0.35 | |
Cow housing | Free stall | Free stall | ||
| Manure handling from cows | ||||
| Collection method | Flush | Scrape | ||
| Storage type | Top-loaded tank | Top-loaded tank | ||
| Storage capacity | Six months | Six months | ||
| Bedding type | Manure solids | Manure solids | ||
| Crop land | ha | 1400 | 1100 | |
| Alfalfa | ha | 280 | 280 | |
| Grass | ha | 50 | 50 | |
| Pasture[b] | ha | 120 | 120 | |
| Corn | ha | 950 | 650 | |
Small grain (double crop) | ha | 470 | 650 | |
| Soil type | Silt loam | Silt loam | ||
Soil phosphorus level | Very high | Very high | ||
| Tillage method | No-till | No-till | ||
| Crop harvest | ||||
| Alfalfa | 5-cut silage | 5-cut silage | ||
| Grass | 2-cut silage | 2-cut silage | ||
| Corn | Silage and grain | Silage | ||
| Small grain | Silage | Silage | ||
| Silage storage | Bunkers | Bunkers | ||
[a] All forge and corn grain fed were produced on the all-feed farm, and grain production was removed from the forage only farm. [b] Pasture used for heifers and nonlactating cows. | ||||
The farm consisted of 2,000 Holstein dairy cows and 1,900 replacement heifers with 1,400 ha of feed-producing cropland (table 1). Essentially all the forage and grain needed to maintain the herd was produced on the farm. Supplemental feeds of canola meal, soybean meal, expeller’s soybean meal, blood meal, distiller’s grain and fat were used to meet protein and energy requirements. Cow rations were formulated to contain 17.5% protein (2.8% N). Phosphorus was fed to meet NRC (2001) recommended requirements for each of six animal groups on the farm (Rotz et al., 2018). Annual milk production was 11,200 kg/cow (10,680 kg/cow corrected to 4% fat and 3.3% protein) with cows milked 3 times per day in a double 40 parlor.
Lactating cows were maintained in mechanically ventilated free stall barns where manure was removed using a flush system with 3 flushes per day. Solids removed from the flushed manure using rotary screens and screw presses were recycled as bedding. Daily bedding use was about 2.3 kg (0.7 kg DM) per cow. Separated water was recycled as flush water. Most, but not necessarily all manure, was separated. Manure beyond that needed to meet the bedding and flush water requirements was returned to cropland. Manure was stored for up to 6 months in various tanks and basins. Heifers were located in different areas of the farm and their manure was handled in slurry and semi-solid forms, depending upon their housing facility.
Feed crops produced on the base farm included alfalfa, grass, corn, and rye with all crops established using a no-till system. Alfalfa was harvested as silage with 5 cuttings per year. Most of the grassland was used as pasture for heifers and dry cows with a smaller portion harvested as silage. About 60% of the corn land was harvested as silage with the remainder harvested as grain. Rye was double cropped on corn silage land, i.e., it was planted in late summer following corn silage harvest and harvested as silage in the spring, prior to establishment of the succeeding crop. Nitrogen fertilizer was applied to corn (140 kg N/ha) and rye (50 kg/ha) to meet crop requirements. All silage was stored in bunker silos.
As an alternative cropping strategy, corn land was reduced to 650 ha with all remaining corn harvested as silage (table 2). Use of the small grain double-crop following corn silage harvest was also set to 650 ha. With more manure applied to corn land, N fertilizer use on corn was reduced to 20 kg N/ha. All other crop and animal management practices remained the same. With all concentrate feeds imported to the farm, a whole-farm imbalance of P was created.
Predominant soils on the farm were Hublersburg and Hagerstown silt loams. Physical characteristics were 53% silt, 21% clay, and 26% sand with a bulk density of 1.3 g/cm3 and water holding capacity of 178 mm. The natural soils were acidic; however, lime was added to the manure solids used for bedding, which has led to high pH levels in heavily manured fields. No additional lime was used on the cropland.
Integrated Farm System Model
The IFSM is a process-level simulation tool used to assess the performance, environmental impacts, and economics of crop, dairy, or beef production systems (USDA-ARS, 2021). The model simulates crop production, feed use, animal production, and the return of manure nutrients back to the land for up to 25 years of weather. Daily growth and development of crops are predicted as a function of soil water and N availability, ambient temperature, and solar radiation (Rotz et al., 2018). Simulated tillage, planting, and harvest operations predict labor, fuel, and other resources used, timeliness of operations, crop losses, and nutritive quality of feeds produced.
Nutrients of N, P, K, and carbon (C) are tracked to predict soil accumulation and losses to the environment (Rotz et al., 2018). Paths of N loss include ammonia (NH3) volatilization, nitrous oxide (N2O), N oxide (NOx), and dinitrogen emissions through nitrification and denitrification processes and leaching and runoff of nitrate (NO3-) and organic or particulate-bound N. Carbon emissions include methane (CH4) from enteric and manure sources along with biogenic carbon dioxide (CO2). Anthropogenic CO2 emissions are those released through fossil fuel combustion and the decomposition of lime and urea fertilizer. Other simulated losses include erosion of sediment and runoff of sediment-bound and dissolved P across the farm boundaries. Emission processes are modeled using dynamic relationships influenced by temperature, wind speed, precipitation, soil conditions, and management practices. Whole-farm mass balances of N, P, K, and C are determined as the sum of nutrient imports in feed, fertilizer, deposition, and fixation, minus the nutrient exports in losses, manure leaving the farm, and feeds, milk, and animals sold.
Following guidelines published by the Livestock Environmental Assessment and Performance partnership (LEAP, 2016), a cradle-to-farm gate life cycle assessment is performed within IFSM to determine annual greenhouse gas (GHG) emissions and fossil energy use (Rotz et al., 2018). Total GHG emission is the sum of emissions of CH4, N2O, and anthropogenic CO2 converted to CO2 equivalents (CO2e) using 100-year global warming potentials of 28 for CH4 and 265 for N2O (Myhre et al., 2013). Emissions include both direct emissions from the production system, as well as indirect N2O emissions that occur elsewhere in the environment resulting from transformation of NH3 and NO3- lost from the farm. Following IPCC (2006) guidelines and recommended factors, portions of these emissions are considered as indirect land emissions of N2O. Our analysis included commodity-based environmental intensities expressed per unit of fat and protein corrected milk (FPCM) adjusted to 4.0% fat and 3.3% protein (IDF, 2015; Rotz et al., 2018).
Greenhouse gas emissions associated with the production of resources used on dairy operations (fuel, electricity, fertilizer, purchased feeds and supplements, purchased animals, machinery, seed, and pesticides) are included in the LCA (Rotz et al., 2018). Fossil energy use includes fuel used in farm operations (machinery and truck use) along with fossil energy used to produce resources consumed on the farm such as electricity, fertilizer, and purchased feeds. Quantities of each used are obtained through the farm simulation and multiplied by appropriate emission or consumption factors for these upstream sources (table 2 in Rotz et al., 2021b). All emissions and resource requirements for handling the manure on the farm were included in our assessment. Emissions from manure or nutrients after leaving the farm were considered external to the production system and were not included.
Simulated performance is used to determine production costs. A whole-farm budget includes fixed and operating costs (Rotz et al., 2018). Annual fixed costs are determined by amortizing the initial cost of equipment or structures over their economic life. Annual operating costs are the simulated resource requirements (fuel, electricity, fertilizer, labor, etc.) times assigned prices. Costs are provided for manure handling along with total production costs. By comparing simulation results, differences among production systems can be determined, including annual resource use, production efficiency, environmental impacts, and production costs.
Previous studies have evaluated and verified IFSM’s accuracy in representing production systems and their environmental impacts. Numerous studies have verified the model’s ability to represent crop production, animal performance, emissions, and other components of dairy and beef cattle production (e.g., Belflower et al., 2012; Rotz et al., 2013, 2014; Bonifacio et al., 2015; Jego et al., 2015; Leytem et al., 2018). The model has been used to evaluate whole-farm environmental impacts and mitigation strategies for dairy farms in the Northeast using recent historical (Veltman et al., 2018) and projected future climate (Veltman et al., 2021). The model has also been used to represent all dairy farms in the U.S. to determine national environmental impacts of the dairy industry (Rotz et al., 2021b).
Modeling of Manure Separation Processes
For this study, processes were added to IFSM to simulate solids and nutrient separation of manure. The initial system on the farm consisted of a rotary screen and screw presses to remove manure solids (fig. 1). Flushing of the free stall barns discharged dirty flush water (about 5% DM) to a small in-ground surge pit. This pit was equipped with two agitators powered by 22-kW motors to keep the solids in suspension. Manure from the surge pit was either pumped for solids removal or sent to a large manure storage using a 15-kW motor. Manure used for solids removal flowed through a rotary screen. Liquid from the screen went to a small holding tank, and the dewatered slurry went to a separate holding tank where two 22-kW agitators kept the solids in suspension. Slurry from this tank was pumped to the screw presses using a 15-kW motor. Two screw presses, each powered by 7.5-kW motors produced manure solids at 30% DM content that were used as bedding material. Liquid from the presses (about 4% DM content) drained by gravity to the same small holding tank used for the rotary screen liquid. A 15-kW pump was used to transfer the liquid to settling lagoons or a large storage tank. Liquid from the settling lagoons provided flush water.
Equipment for solids separation and flushing was added as a model component in IFSM. Assumed parameters included the initial cost, annual repair and maintenance cost, and throughput capacity of the equipment (table 2). Daily and annual use of the equipment were based upon the amount of manure processed and throughput capacity of the equipment. Total power required to operate the system was the sum of all motors used. Electricity use was determined assuming a 70% load on motors times their hours of use. Labor for operating this equipment was small and was set at 1% of the annual operating time of the equipment.
 |
| Figure 1. Flush manure handling system using rotary screens and screw presses to remove 80% of manure solids, which are used for bedding. |
The centrifuge system included all the equipment and processes used for solids separation with a centrifuge added prior to the liquid flowing to the storage lagoons (fig. 2). A refurbished centrifuge was assumed with an initial cost of $150,000, which included new controls and plumbing and electrical installation costs (table 2). The throughput capacity of the centrifuge was 2.52 L/s or 9.3 t/h. We assumed the centrifuge would operate 17 h per day, which processed about 60% of the manure effluent from the screw presses. Based on information from a centrifuge supplier, we estimated annual repair and maintenance costs for the centrifuge and pumps to be $2.70 per hour of use. The centrifuge removed an additional 2% of the initial manure solids and 25% of the P from the portion of the manure flowing through the centrifuge (Church et al., 2018). The centrifuge used two motors with a total power rating of 41 kW. Feed and effluent pumps were also used requiring 7.5 kW each. The centrifuge was placed inside a structure to prevent freezing when not in operation (table 2). Operator labor for this equipment was small and assumed to be 1% of the annual operating time.
| Table 2. Characteristics of the manure handling systems used on simulated dairy farms in Pennsylvania. | ||||||||||
| All-feed Farm, Flush System | All-feed Farm, Scrape System | Forage Farm, Scrape System | ||||||||
| System Component | Units | Separator[a] | Separator & Centrifuge | Separator[a] | Separator & Centrifuge | Separator[a] | Separator & Centrifuge | |||
| Rotary screen and screw press[a] | ||||||||||
| Initial equipment cost | $ | 250,000 | 250,000 | 125,000 | 125,000 | 125,000 | 125,000 | |||
| Repair & maintenance | $/h of use | 10 | 10 | 5 | 5 | 5 | 5 | |||
| Annual use[b] | h | 3,402 | 3,366 | 3,401 | 3,364 | 3,419 | 3,380 | |||
| Power requirement | kW | 125 | 125 | 88 | 88 | 88 | 88 | |||
| Labor requirement | % of annual use | 1 | 1 | 1 | 1 | 1 | 1 | |||
| Throughput capacity | t/h | 23.0 | 23.0 | 11.5 | 11.5 | 11.5 | 11.5 | |||
| Centrifuge[c] | ||||||||||
| Initial equipment cost | $ | 150,000 | 150,000 | 150,000 | ||||||
| Initial facility cost | $ | 20,000 | 20,000 | 20,000 | ||||||
| Repair & maintenance | $/h of use | 2.70 | 2.70 | 2.70 | ||||||
| Annual use[b] | h | 6,132 | 3,066 | 2,993 | ||||||
| Power requirement | kW | 56 | 56 | 56 | ||||||
| Labor requirement | % of annual use | 1 | 1 | 1 | ||||||
| Throughput capacity | t/h | 9.3 | 9.3 | 9.3 | ||||||
| Manure hauling | ||||||||||
| Initial truck cost | $/truck | 57,500 | 57,500 | 57,500 | 57,500 | 57,500 | 57,500 | |||
| Repair & maintenance[b] | $/truck/yr | 5,376 | 6,042 | 6,924 | 8,160 | 6,140 | 6,487 | |||
| Number of trucks | 7 | 5 | 6 | 4 | 5 | 5 | ||||
| Average distance | km | 7.0 | 3.5 | 7.0 | 3.5 | 4.0 | 4.0 | |||
| Hauling time[b] | truck-h | 5,163 | 2,650 | 5,294 | 2,314 | 3,136 | 3,289 | |||
[a] With flush system, the separator includes a rotary screen and two screw presses to remove 80% of the solids of treated manure. Scrape systems use no rotary screen and one screw press. [b] Values determined through 25-year simulations of the alternative manure handling systems. [c] Centrifuge removes 1.5% of the original manure solids and up to 32% of the manure P entering the device. | ||||||||||
 |
| Figure 2. Flush manure handling system using rotary screens, screw presses and a centrifuge to remove 82% of the manure solids. The centrifuge removes 25% of the phosphorus in the manure processed by the device. |
A flush system requires the handling of large volumes of material. The farm represented in our assessment flushes 3 times per day using 114,000 L/flush. This more than doubles the amount of material that must move through the separation equipment. Although this large volume of water is being recycled, handling this volume greatly increases operating time, wear on the equipment, and electricity use. Another option is to scrape the manure from the barn floor at a dry matter content of 12% to 14% compared to 5% to 6% in flush water. This alternative handling system is illustrated in figure 3. With the more concentrated manure, rotary screens are no longer needed to thicken the manure and one screw press is able to handle the required volume of manure (table 2). The centrifuge operates more efficiently extracting 32% of the P flowing through the device. All other characteristics of the manure system remain the same as those of the base flush system. These changes reduce the initial investment in equipment and the repair and maintenance costs by 50%. Throughput capacity is also halved giving a similar annual operating time with the lower volume of material processed.
 |
| Figure 3. Scrape manure handling system using a screw press and centrifuge to remove 82% of the manure solids. The centrifuge removes 32% of the phosphorus in the manure processed by the device. |
Costs for owning and operating the manure handling systems were determined using the economic component in IFSM (Rotz et al., 2018). All equipment and facilities were amortized over an economic life and the annualized cost was added to other operating costs to get a total production cost. Manure handling cost included the fixed and operating (repair and maintenance, fuel, and labor) costs of the rotary screen, screw presses, and centrifuge. Manure hauling was also an important cost in the assessment because the number of trucks required and hauling distance varied among systems (table 2). Hauling cost included the amortized initial cost of trucks, annual costs for repair and maintenance, fuel, and operator labor. These costs were a function of the carrying capacity of the trucks (19 t per load), loading time (7-9 min), travel speed (8 km/h during application, 40 km/h road speed), and average distance traveled (table 2).
Simulated Manure Systems
Six production systems were evaluated through IFSM simulations (table 2). Separation strategies were only applied to manure from the cow barns; manure systems for heifer barns remained unchanged. The first two systems used the current farm where most of the feed was produced on the farm with no excess P in the production system. The issue for this system was one of distribution where some of the manure must be hauled long distances to return the manure P to outlying fields. Removing and concentrating the P provided a material that could be hauled and distributed more efficiently. Use of the manure flush system created a large volume of dilute solution flowing through the separation equipment.
For the next two systems, a scrape manure system was used to remove manure from the barns in place of the flush system. This reduced the volume of manure passed through the screw press and centrifuge. With this reduced volume, the rotary screen was not needed, only one screw press was required, and the operating time of the centrifuge was reduced. This reduced the total power required to operate the equipment and the electricity consumed. Manure was scraped from the free stall barn floor each day while cows were being milked using a skid-steer loader. Fuel use, labor, and other costs of the added operation were included in the modeling of this system. The thicker slurry manure was cycled through the screw press to provide solids for bedding. The primary benefit of this system was that the operating time of the separation equipment was reduced by over half by processing the less dilute solution at the fixed flow rates of the equipment. The volume of manure transported and applied to cropland was a little less than that of the flush system, so the number of trucks and hauling requirements were reduced (table 2).
For the final two simulations, the land area was reduced by 31%, and no corn grain was produced. With all corn harvested as silage, the entire corn land could be double cropped with winter small grain crops. For this scenario, all the forage required for the herd was produced on-farm and all concentrate feed was purchased and imported to the farm. This created an excess of P beyond that taken up by the feed crops. Thus, to maintain a long-term farm balance, some of the P was removed from the farm. Extracting the manure P provided a product that was more efficiently transported to distant farms or other enterprises that needed the P. The goal for the analysis of this scenario was to determine the cost of extracting the P expressed per unit of P obtained. This represented the cost of removing the excess P from the farm to avoid long-term accumulation in the soil of the farm. This cost could be compared to the price of commercial fertilizer to obtain a value for the product relative to the alternative of purchasing commercial fertilizer.
Important economic costs of manure handling included the costs of owning, maintaining, and operating the rotary screen, screw presses, centrifuge, and trucks for transporting and field applying the manure. Initial costs of the equipment varied with the number of rotary screens and screw presses required and the number of trucks required varied with the hauling distance and volume of manure applied to cropland (table 2). More trucks were required to handle all manure without P extraction where the average hauling distance was 7 km. With the P extracted, the manure did not need to be transported as far to avoid P accumulation in the land near the barns. Thus, fewer trucks were required to haul the manure an average distance of 3.5 km (table 2).
Initial investments in equipment were amortized over their economic life (normally 10 years) using a real annual discount rate of 2% with an end-of-life value of 30% of the initial cost. Structures were amortized over 25 years with no end-of-life value. Repair and maintenance costs varied with annual use and the initial cost of manure handling equipment. Important prices for this economic analysis were that of fuel ($0.85/L), electricity ($0.10/kWh), N fertilizer ($1.32/kg N), and labor ($12/h).
Sensitivity Analysis
A sensitivity analysis was conducted to evaluate the relative influence assumed parameters had on the economics of the extraction process. This analysis identifies components that have the greatest effect on the results, or it indicates the associated error if an incorrect assumption affected this component. A sensitivity index was determined as the ratio of the percent change in the assessed output over a 20% change in the tested input. For this analysis, the output was defined as the economic benefit or loss obtained with the use of the P extraction process. A sensitivity index near 0 indicates low sensitivity and hence, minor change in the predicted output resulting from a change in the component, whereas indices close to or greater than 1 indicate high sensitivity. With high sensitivity higher data quality are needed for these parameters for accurate simulation. Sensitivity indices were determined for initial cost, repair and maintenance cost, electricity use, labor requirement, and throughput capacity of the centrifuge; initial, repair and maintenance, fuel, and labor costs of trucks; and transport distance.
Results and Discussion
Use of a centrifuge to extract P from dairy manure was evaluated using the three production systems with and without extraction. The first production system was the current farm with manure collected by flushing, next was the same farm with manure collection by scraping and the last was a modified farm with scraped manure where only forage crops were produced and concentrate feeds were purchased.
Farm with Grain Production and Flush Manure System
A problem for the current farm was that a portion of the cropland was not near the cow housing facility where manure is produced. Extraction of the P provided a material with concentrated P for more economical hauling. Manure liquid in excess of that needed for flush water can be applied closer to the animal facility. This solution with a greater N to P ratio meets the needs of those crops. Inorganic N fertilizer that is currently purchased can be used along with the concentrated P material to meet the needs of cropland more distant from the manure source. To represent this approach, hauling distance was reduced by 50% and the number of trucks was reduced by 29%. Implementation of this approach reduced manure hauling time and truck repair and maintenance costs by 25% (table 2). Annual fuel use in manure handling was reduced 16%, but this was offset by a 79% increase in electricity use. Together, energy costs increased 35% (table 3). Annual costs of owning and maintaining equipment and facilities increased with the flush system, but labor cost was reduced through less manure hauling time (table 3). With all economic changes considered, annual manure handling costs increased by $27,674. This indicates that the use of the centrifuge for P extraction was not economically justified for this system where large volumes of low concentration liquid manure must be handled.
Use of the P extraction process did not have much influence on the environmental impacts of the farm (table 3). Phosphorus runoff losses were reduced slightly with the redistribution of P. Because all P remained on the farm, a long-term P balance was maintained with no accumulation in the soil. The farm-gate life cycle intensity of fossil energy use increased by 3% through increased electricity use offset by reduced fuel use. This resulted in less than a 1% increase in the intensity of GHG emissions.
Farm with Grain Production and Scrape System
Shifting from the flush system to scraping of manure greatly reduced electricity consumption with little change in fuel consumption (table 3). Electricity use was reduced due to removal of the rotary screen and the lower volume of manure processed through the screw press. With one less screw press use, the operating time was similar to that found for the flush system (table 2). Energy costs were reduced 24%, but labor cost increased 65% (table 3). Overall, manure handling costs were reduced 9% by changing the flush system to a scrape system providing more economical manure handling.
| Table 3. Environmental impacts and manure handling costs of simulated dairy farms without and with a centrifuge to remove manure nutrients. | ||||||||||||||||||||
| All-Feed Farm, Flush System | All-Feed Farm, Scrape System | Forage Farm, Scrape System | ||||||||||||||||||
Impact or Cost Category | Separator | Separator & Centrifuge | Separator | Separator & Centrifuge | Separator | Separator & Centrifuge | ||||||||||||||
| Energy use | ||||||||||||||||||||
| Electricity | kWh/yr | 326,303 | 585,287 | 179,958 | 309,171 | 180,891 | 306,916 | |||||||||||||
| Fuel | L/yr | 33,363 | 28,049 | 33,207 | 27,342 | 28,331 | 24,128 | |||||||||||||
| Nitrogen loss | ||||||||||||||||||||
| Volatilized | kg/ha | 66.1 | 65.5 | 97.6 | 97.0 | 123.0 | 121.3 | |||||||||||||
| Denitrified | kg/ha | 30.9 | 30.5 | 11.9 | 11.8 | 13.1 | 11.8 | |||||||||||||
| Leached | kg/ha | 69.2 | 69.1 | 63.4 | 62.8 | 32.0 | 29.7 | |||||||||||||
| Runoff | g/ha | 1,370 | 1,376 | 1,339 | 1,329 | 974 | 945 | |||||||||||||
| Phosphorus | ||||||||||||||||||||
| Soluble runoff | g/ha | 194 | 183 | 225 | 211 | 153 | 151 | |||||||||||||
| Sediment runoff | g/ha | 1,611 | 1,603 | 1,602 | 1,592 | 1,255 | 1,254 | |||||||||||||
| Soil accumulation | kg/ha | 0 | 0 | 0 | 0 | 6.8 | 0 | |||||||||||||
| P extracted | kg/year | 6,000 | 6,000 | 7,480 | ||||||||||||||||
| Life cycle environmental intensities | ||||||||||||||||||||
| Fossil energy use | MJ/kg FPCM[a] | 2.51 | 2.59 | 2.47 | 2.51 | 2.36 | 2.43 | |||||||||||||
| Greenhouse gas | kg CO2e/kg FPCM | 0.88 | 0.88 | 0.88 | 0.88 | 0.87 | 0.87 | |||||||||||||
| Annual manure handling costs | ||||||||||||||||||||
| Equipment | $ | 155,216 | 164,314 | 128,448 | 126,838 | 119,772 | 127,703 | |||||||||||||
| Facility | $ | 28,534 | 29,657 | 28,534 | 29,657 | 28,534 | 29,657 | |||||||||||||
| Energy | $ | 61,089 | 82,455 | 46,321 | 54,240 | 42,255 | 51,273 | |||||||||||||
| Labor | $ | 25,534 | 21,621 | 42,244 | 38,106 | 37,708 | 38,422 | |||||||||||||
| Total | $ | 270,373 | 298, 047 | 245,547 | 248,841 | 228,269 | 247,055 | |||||||||||||
| Net extraction cost | $ | 27,674 | -2,331 | 18,786 | ||||||||||||||||
| Net extraction cost | $/kg P | 4.61 | -0.39 | 2.51 | ||||||||||||||||
[a] FPCM is milk corrected to 4% fat and 3.3% protein. | ||||||||||||||||||||
The change to a scrape system had mixed effects on the environment (table 3). Ammonia emissions from the barn floor and manure storage increased through the handling of slurry manure in the scrape system (Rotz et al., 2014). With greater ammonia emission prior to field application, less N was lost from the cropland through nitrification, denitrification, leaching and runoff processes. Due to the greater dry matter content, less of the manure infiltrated into the soil creating a slight increase in soluble P runoff while sediment bound P decreased. The farm-gate life cycle energy consumption decreased 2% with no change in greenhouse gas emission intensity.
Phosphorus extraction with the centrifuge was more efficient and economical when used with the scrape system (table 3). Compared to the base scrape system, addition of the centrifuge increased electricity consumption and reduced fuel consumption similar to that found with the flush system. As found with the flush system, use of the centrifuge had little effect on the environment other than a 2% increase in the farm-gate intensity of fossil energy consumption. Equipment and labor costs were reduced with the use of the centrifuge through reduced manure hauling distance. Facility and energy costs increased, but the total annual cost of manure handling decreased by $2,331. Thus, with the use of the scrape manure system, the benefit received through reduced manure handling more than offset the increased costs of owning and operating the centrifuge.
Farm with Only Forage Production and Scrape System
Use of a farm that only produced forage to feed the herd enabled an assessment of the removal and export of excess P. Compared to the farm with the flush system, less electricity was used, and with less hauling distance for the manure, less fuel was used (table 3). Because the volume of manure handled was similar to that of the larger farm using the scrape system, electricity consumption was similar. With a smaller land base, hauling distance was less and fewer trucks were required to spread the manure. This reduced the amortized equipment costs and total manure handling costs compared to the previous farm systems (table 3).
Volatile N losses were similar to those using the scrape system on the farm with the larger crop area. Losses expressed per hectare were greater though when divided by the smaller land base (table 3). With less cropland, manure N supplied a greater portion of the N required for crop growth and less inorganic fertilizer was used. These changes led to less leaching and runoff loss of N. With less corn land on the farm and all the corn double cropped with small grains, P runoff losses expressed per hectare were reduced by 22% compared to the other farm systems (table 3). Farm-gate fossil energy use and greenhouse gas emissions were also slightly less than that found for the prior farm systems.
To remove excess P from the farm (7,480 kg P/yr), the centrifuge was operated 8.2 h/d. Centrifuge use increased amortized equipment and energy costs with small increases in facility and labor costs. This increased the annual manure handling cost by $18,786 compared to the same farm without the centrifuge. This added cost was $2.51/kg of P in the solids leaving the centrifuge. Compared to phosphate fertilizer prices ($1.50 to $2.00/kg P), marketing this material for its P value alone may not be economically viable. Preliminary data from experimental tests using a centrifuge on the farm modeled in this study has shown that the centrifuge also extracts substantial portions of the calcium (53%), iron (33%), manganese (28%), and magnesium (17%) in the manure along with small amounts of N, K, and other micronutrients (Clinton Church, USDA-ARS, Unpublished data, 20 January 2022). Considering the value of all nutrients would improve the economic value of the extracted product, but the amounts and economic value of these nutrients has not been explicitly quantified. The wet material extracted must be used locally and soon after extraction. Otherwise, the added cost of drying and further processing would be required to stabilize and market a product for broader use. A specialized market, such as organic, may also increase the market value of the product.
The primary environmental benefit for using centrifuge extraction on this farm was the removal of excess P to prevent long-term accumulation in the farm soil. With little soil accumulation, future losses of P in runoff should be reduced. Removal of this P had little effect on present or near-term runoff losses (table 3), but the long-term risk of eutrophication of surface waters is reduced. Although this reduced risk has little direct economic benefit to the producer, it offers less tangible, long-term benefit to society. As found with the other farm systems, use of the centrifuge extraction process provided minor reductions in N losses and a small increase in farm-gate fossil energy use with little effect on GHG emissions (table 3).
Sensitivity Analysis
For farms that produce most of their feed, the benefit of P extraction comes through reduced transport distance for the manure. With a flush system, the increase in manure handling cost using the centrifuge is most sensitive to the throughput capacity and amount of electricity consumed by the centrifuge (table 4). With a sensitivity index around 1, a 20% increase in throughput capacity or reduction in electricity use reduces the net increase in manure handling cost about 20%. This increased cost is moderately sensitive to the initial cost and the repair and maintenance cost of the centrifuge. Labor required to operate the centrifuge is small with little effect on manure handling cost. For this strategy, the increased manure handling cost is also moderately sensitive to truck transport costs (table 4). The economic loss is not sensitive to the manure transport distance when distances without and with the extraction process are reduced together. When only the distance using the extraction process is reduced, the economic loss is reduced but remains a loss. Relatively large reductions in these components are needed to provide economically viable use of the centrifuge with the flush system.
When the centrifuge is used with a scrape manure system, the economic value of the extraction process is very sensitive to most of the operating parameters assumed for both the centrifuge and truck transport (table 4). The greater sensitivity is due to the small difference in manure handling costs between systems without and with the use of the centrifuge. A 20% increase in throughput capacity or 20% reduction in the initial cost, repair and maintenance cost, or electricity use of the centrifuge provides more than a 70% increase in the economic benefit obtained using the centrifuge. Similar increases in economic benefit were obtained by reducing the truck initial, repair and maintenance, fuel, and labor costs or transport distance (table 4).
| Table 4. Sensitivity of the economic value of the P extraction process to assumed parameters for owning and operating the centrifuge and manure transport trucks.[a] | ||||||
| All Feed Farm | Forage Only Farm | |||||
| System Component | Flush | Scrape | Scrape | |||
| Centrifuge | ||||||
| Initial equipment cost | 0.48 | 5.7 | 0.71 | |||
| Repair & maintenance | 0.60 | 3.6 | 0.43 | |||
| Electricity use or cost | 0.95 | 5.6 | 0.68 | |||
| Throughput capacity | 1.04 | 4.4 | 0.73 | |||
| Labor requirement or cost | 0.03 | 0.2 | 0.02 | |||
| Truck transport | ||||||
| Initial cost | 0.37 | 4.4 | 0.08 | |||
| Repair & maintenance | 0.37 | 3.8 | 0.01 | |||
| Fuel use or cost | 0.16 | 2.2 | 0.19 | |||
| Labor cost | 0.01 | 2.4 | 0.01 | |||
| Transport distance | 0.85 | 2.5 | 0.01 | |||
| Extracted P | 0.83 | |||||
[a] Sensitivity index is the change in manure handling cost created by use of the centrifuge divided by a 20% change in the modeled component. An index less than 0.2 is considered low sensitivity, 0.2 to 0.6 is moderate sensitivity and greater than 0.6 is high sensitivity. | ||||||
For farms where excess P is removed to provide a long-term balance, the economic cost of extracting that P is moderately sensitive to the repair and maintenance cost, and highly sensitive to the initial cost, electricity use or cost, and throughput capacity of the centrifuge (table 4). A 20% change of each reduces the net increase in manure handling costs by 9% to 15%. Because the export of this small amount of manure solids has little effect on manure handling, changes in trucking costs have little effect on the economics of P extraction within this scenario. When expressed per unit of extracted P, the extraction cost is highly sensitive to the amount of P removed with a 20% improvement providing a 17% reduction in the increased cost. A 20% improvement in these more highly sensitive parameters will bring the extraction cost closer, but not less than, the price of inorganic phosphate fertilizer.
Conclusions
Use of a centrifuge for P extraction from dairy manure appears less feasible and uneconomical when a flush system is used for barn removal of manure due to the large volume of dilute material that must be handled. When barn floors are scraped, the volume of material processed is reduced, which improves extraction efficiency and reduces electricity consumption providing more economical P extraction.
On a large dairy farm where manure must be transported long distances to obtain uniform distribution, P extraction with a centrifuge provides a better ratio of N and P contents for use on nearby cropland and reduces transport costs for nutrients applied to more distant cropland. When used with a scrape manure system, increased extraction costs may be more than offset by reduced transport costs.
To avoid long-term accumulation of soil P on farms where substantial portions of feed are imported, use of centrifuge extraction provides a material with a high P concentration that can be exported from the farm for other uses. The cost of producing this P rich material may be greater than that of purchased phosphate fertilizer prices. Although marketing this material for its P content alone may not be economically beneficial to the dairy producer, the material may have additional value considering other nutrients and characteristics and the reduction in long-term risk of surface water eutrophication benefits society.
Acknowledgements
This work was supported by the U.S. Department of Agriculture, Agricultural Research Service. The authors thank the producer who contributed characteristics of their farm.
References
Belflower, J. B., Bernard, J. K., Gattie, D. K., Hancock, D. W., Risse, L. M., & Rotz, C. A. (2012). A case study of the potential environmental impacts of different dairy production systems in Georgia. Agric. Syst., 108, 84-93. https://doi.org/10.1016/j.agsy.2012.01.005
Bonifacio, H. F., Rotz, C. A., Leytem, A. B., Waldrip, H. M., & Todd, R. W. (2015). Process-based modeling of ammonia and nitrous oxide emissions from open-lot beef and dairy facilities. Trans. ASABE, 58(3), 827-846. https://doi.org/10.13031/trans.58.11112
Church, C. D., Hristov, A. N., Bryant, R. B., Fishel, S. K., & Kleinman, P. J. (2016). A novel treatment system to remove phosphorus from liquid manure. Appl. Eng. Agric., 32(1), 103-112. https://doi.org/10.13031/aea.32.10999
Church, C. D., Hristov, A. N., Kleinman, P. J., Fishel, S. K., Reiner, M. R., & Bryant, R. B. (2018). Versatility of the manure phosphorus extraction (MAPHEX) system in removing phosphorus, odor, microbes, and alkalinity from dairy manures: A four-farm case study. Appl. Eng. Agric., 34(3), 567-572. https://doi.org/10.13031/aea.12632
Church, C. D., Hristov, A., Bryant, R. B., & Kleinman, P. J. (2017). Processes and treatment systems for treating high phosphorus containing fluids. U.S. No. Patent 10,737958.
IDF. (2015). A common carbon footprint approach for dairy. Bull. 479/2015. Brussels, Belgium: International Dairy Federation. Retrieved from www.fil-idf.org/wp-content/uploads/2016/09/Bulletin479-2015_A-common-carbon-footprint-approach-for-the-dairy-sector.CAT.pdf
IPCC. (2006). Guidelines for national greenhouse inventories. Vol. 4: Agriculture, forestry and other land use. International Panel on Climate Change. Retrieved from www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.html
Jego, G., Rotz, C. A., Belanger, G., Tremblay, G. F., Charbonneau, E., & Pellerin, D. (2015). Simulating forage crop production in a northern climate with the Integrated Farm System Model. Can. J. Plant. Sci., 95(4), 745-757. https://doi.org/10.4141/cjps-2014-375
Kleinman, P., Blunk, K. S., Bryant, R., Saporito, L., Beegle, D., Czymmek, K.,... Smith, M. (2012). Managing manure for sustainable livestock production in the Chesapeake Bay Watershed. JSWC, 67(2), 54A-61A. https://doi.org/10.2489/jswc.67.2.54A
LEAP. (2016). Environmental performance of large ruminant supply chains: Guidelines for assessment. Rome, Italy: United Nations FAO. Retrieved from www.fao.org/3/a-i6494e.pdf
Leytem, A. B., Bjorneberg, D. L., Rotz, C. A., Moraes, L. E., Kebreab, E., & Dungan, R. S. (2018). Ammonia emissions from dairy lagoons in the Western U.S. Trans. ASABE, 61(3), 1001-1015. https://doi.org/10.13031/trans.12646
Myhre, G., Shindell, D., Breon, F.-M., Collins, W., Fuglestvedt, J., Huang, J.,... Zhang, H. (2013). Anthropogenic and natural radiative forcing. In T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung,... P. M. Midgley (Eds.), Climate Change 2013: The physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U. K., New York, NY: Cambridge University Press.
NRC. (2001). Nutrient requirements of dairy cattle (7th rev. ed.). Washington, DC: National Research Council: Natl. Acad. Sci.
Pagliari, P. H., & Laboski, C. A. M. (2012). Investigation of the inorganic and organic phosphorus forms in animal manure. JEQ, 41(3), 901-910. https://doi.org/10.2134/jeq2011.0451
Pagliari, P. H., & Laboski, C. A. M. (2013). Dairy manure treatment effects on manure phosphorus fractionation and changes in soil test phosphorus. Biol. Fertil. Soils, 49(8), 987-999. https://doi.org/10.1007/s00374-013-0798-2
Rotz, C. A., Stout, R., Leytem, A., Feyereisen, G., Waldrip, H., Thoma, G.,... Kleinman, P. (2021b). Environmental assessment of United States dairy farms. J. Cleaner Prod., 315, 128153. https://doi.org/10.1016/j.jclepro.2021.128153
Rotz, C. A., Asem-Hiablie, S., Cortus, E. L., Spiehs, M. J., Rahman, S., & Stoner, A. M. (2021a). An environmental assessment of cattle manure and urea fertilizer treatments for corn production in the Northern Great Plains. Trans. ASABE, 64(4), 1185-1196. https://doi.org/10.13031/trans.14275
Rotz, C. A., Corson, M. S., Chianese, D., Montes, F., Hafner, S. D., Bonifacio, H. F., & Coiner, C. U. (2018). Integrated farm system model: Reference manual. University Park, PA: USDA-ARS. https://www.ars.usda.gov/ARSUserFiles/80700500/Reference%20Manual.pdf
Rotz, C. A., Isenberg, B. J., Stackhouse-Lawson, K. R., & Pollak, E. J. (2013). A simulation-based approach for evaluating and comparing the environmental footprints of beef production systems. J. Anim. Sci., 91(11), 5427-5437. https://doi.org/10.2527/jas.2013-6506
Rotz, C. A., Montes, F., Hafner, S. D., Heber, A. J., & Grant, R. H. (2014). Ammonia emission model for whole farm evaluation of dairy production systems. JEQ, 43(4), 1143-1158. https://doi.org/10.2134/jeq2013.04.0121
Sharpley, A. N., Chapra, S. C., Wedepohl, R., Sims, J. T., Daniel, T. C., & Reddy, K. R. (1994). Managing agricultural phosphorus for protection of surface waters: Issues and options. JEQ, 23(3), 437-451. https://doi.org/10.2134/jeq1994.00472425002300030006x
USDA-ARS. (2021). The integrated farm system model, ver. 4.6. University Park, PA: USDA-ARS. Retrieved from https://www.ars.usda.gov/northeast-area/up-pa/pswmru/docs/integrated-farm-system-model
Veltman, K., Rotz, C. A., Chase, L., Cooper, J., Forest, C. E., Ingraham, P. A.,... Jolliet, O. (2021). Assessing and reducing the environmental impact of dairy production systems in the northern U.S. in a changing climate. Agric. Syst., 192, 103170. https://doi.org/10.1016/j.agsy.2021.103170
Veltman, K., Rotz, C. A., Chase, L., Cooper, J., Ingraham, P., Izaurralde, R. C.,... Jolliet, O. (2018). A quantitative assessment of Beneficial Management Practices to reduce carbon and reactive nitrogen footprints and phosphorus losses on dairy farms in the U.S. Great Lakes region. Agric. Syst., 166, 10-25. https://doi.org/10.1016/j.agsy.2018.07.005