Cover Image
close this book Pig waste management and recycling: The Singapore experience
close this folder 5. Environmental impact of pig wastes
View the document Properties of pig wastewater
View the document Environmental impact of the Ponggol PFA
View the document Questions and answers
View the document Forum

Properties of pig wastewater

In the design of waste management and pollution control systems and in the assessment of the environmental impact of waste disposal, it is necessary to quantify waste characteristics in engineering terms. Many terms may be used as waste and water quality parameters. Not all of these parameters need to be quantified before devising systems for waste management. However, the design parameters selected must be those that relate to the process being considered and to the water quality standards to be achieved. Environmental impact assessment includes, besides the effects of wastes on water and soil resources, nuisance factors and social and political reactions that cannot always be quantified.

In confined commercial pig production, as practiced in Singapore and in the tropics in general, wastes consist of feces and urine, wasted feed, spilled water from the drinking nipples and leaking pipes, and water used for cleaning the pens and cooling the pigs. Because wastewaters are collected in open drains, they contain runoff water plus soil and other particles carried by the runoff water. The wastewater generated on farms in the tropics is, therefore, not the same as the manure or slurry of farms in temperate and cold climates.

A sampling program was carried out over several years to arrive at design guidelines for the physical, chemical, and biological properties of the pig wastewaters.

Sampling

The results of any test procedure, no matter how well followed, are no better than the sample taken. Therefore, staff were first trained to take representative samples using integrated and composite sampling procedures and equipment. Sampling was carried out for the duration of the flow or, in some instances, continuously for 24 h.

At first, existing government laboratories were used. However, their limited capacity and flexibility could not cope with the 31 different tests run on replicate samples from the 27 sampling points plus additional samples from experimental programs. Purchase of equipment and minor renovations were sufficient to establish a waste analysis laboratory to complement the existing wet chemistry laboratory that was used for feed analysis and animal nutrition at PPRTI. The same laboratory personnel were used after training.

Sampling Network

A network of sampling points was established for the Ponggol PFA (see Fig. 5.1). Additional surveys in the estuaries and beaches at the outlet of the Serangoon River were carried out by government agencies.

Collection of Samples

Samples were collected in 4-L, wide-mouth plastic bottles from the top layer of flow of the rivers. No bottom mud was collected, except in small drains where the entire flow was sampled. Each sample bottle was identified by type and location. The temperature of the sample and the time of collection were clearly marked. All stream-water sample bottles for dissolved oxygen measurements were completely filled to avoid aeration during transport.

Grab Samples

All stream samples in the main body of the rivers were grab samples. They were taken during low tide, from the top 15 cm of the depth of flow.

Composite Samples

Composite samples were taken over periods of 4-8h. When the construction of a central treatment plant was considered, 24-h composite samples were taken for a week at a time at the proposed site (sampling point 19; see Fig. 5.1).

Sample Storage

All samples were stored in ice boxes on the way to PPRTI, where they were kept at 4°C until analysis. Analysis was initiated within 24 h of collection.

Testing of Samples

The testing procedures for the common parameters used in the characterization of animal wastes and wastewaters were based on The Standard Methods for the Examination of Water, Wastewater, and Sludges as revised every 5 years by the American Public Health Association (Washington, DC) and other cooperating scientific societies.


Fig. 5.1. Location of WWS 19 and typical small farm layout: (A) the Ponggol PFA and its pollution survey network, showing sampling points 1 to 20 (except points 11 and 12 which were further up the Serangoon River); (B) WWS area for sampling point 19; (C) typical farmlot layout.

Waste-Generation Rates

Several approaches were taken to obtain waste-generation rates. They were developed mostly by actual collection of feces and urine in metabolic crates and by monitoring, under controlled conditions, the quantities of waste added to pits below slatted floors and taken from barns with solid concrete floors. The volume of waste was related to the total live weight of the pigs that generated it by weighing the entire pig population during the sampling period.

In one experiment in a building full of porker pigs, the major inputs (feed and water) and outputs (solid and liquid wastes) were measured over several weeks. Solids and water were collected from under the slatted floors in a pit that had been calibrated so that changes in water level were converted to volume of wastewater. The water inputs were measured with water meters.

Water Use

In the aforementioned experiment, water used in the slatted floor barn for drinking and washing was measured for 47 days. During this period, the mean live weight of the porkers was 59.7 kg. The average rate of water use was 17.5 L/porker per day, which was equivalent to 29.3 L/APU, where APU (animal population unit) is 100 kg live weight.

As the pigs grew, daily water usage increased from 13.9 to 23.8 L/porker (27 to 35 L/APU). These figures were combined with those from other pigs to obtain an estimated value for water use for drinking and for cleaning the pens of 20 L/SPP or 36 L/APU. Hourly water withdrawal from the pipeline averaged 0.6 L/APU at night and a little over 2 L/APU during the day. From 0800 to 1700, the porkers consumed 68% of the total water used per day. The water supply was kept under low pressure (less than 3 m) to minimize spillage.

Wastewater Generation

Based on analyses of data over approximately 6 years, design values for both waste and wastewater generation were adapted (Table 5.1). These values were used in the design of all the wastewater-treatment facilities built in Singapore. Table 5.2 compares the Singapore values with major sources of similar data.

Flow Pattern

The wastewater flow pattern of individual farms had a double peak (one in the morning and another in the afternoon) that corresponded to the times at which the pens were hosed down. The flow pattern from an area that served 50 farms (some of which hosed only once) produced a single peak at midday because of overlap and the distance to the sampling point.

The peak flow rate was four times larger than the average flow rate. Generally, the same ratio of four applied to the peak BOD5 compared with the average BOD5. The ratio of the peak BOD5 to the minimum BOD5 ranged from 7:1 to 12:1. Minimum BOD5 values were encountered at night (from 1900 to 0700). These extreme variations are critical in the design of treatment processes that are sensitive to changes in the characteristics of the effluent.

Table 5.1. Properties of pig wastes and wastewaters in Singapore farms.a

 

Parameter b

kg/day per APU

% TWW

% TTS

% TVS

kg/day per SPP

mg/L

TWW (feces + urine)

8.4

-

-

-

4.54

-

TTS

0.84

10

-

-

0.45

18300

TVS

0.67

8

80

-

0.36

14700

TFS (ash; minerals)

0.17

2

20

26

0.09

3700

TSS

0.69

8.2

82

103

0.37

15000

TDS

0.15

2

18

22

0.08

3300

BOD5

0.25

3.0

30

37

0.13

5300

COD

0.84

10

100

125

0.45

18300

TKN

0.05

0.6

6

7.5

0.03

1100

TPO (= 2.27 TPP)

0.02

0.25

2.5

3.0

0.01

440

TKO (= 1.21 TKK)

0.01

0.14

1.4

1.8

0.005

220

Hosing/cooling waters

37

-

-

-

20

-

TWF c

45.4

-

-

-

24.5

-

a These values represent the quantities generated and do not account for any losses after excretion. As such, they represent maximum mean design values for preliminary calculations. I APU (animal population unit) consists of 100 kg total live weight (TLW). SPP (standing pig population) averaged 54 kg for all pigs (sows, piglets, boars, weaners, and porkers). The term mg/L is the concentration of waste expected under the traditional use of water to hose down the pig pens at a daily rate of 20 L/SPP. The total wastewater flow (TWF) = 20 + 4.54 = 24.54 L/SPP per day. Concentrations change in proportion to the volume of water used to remove the wastes each day. In flushing systems, the daily volume of water used should be about 30 L/SPP, making TWF = 30 + 4.54 = 34.54 L/SPP per day.

b See list of Acronyms and Abbreviations for definitions.

c Units: L/day per APU or per SPP.

Cesspit Effects

The farms in Phases I and II of the Ponggol PFA were required by the government to install a cesspit, which was an underground concrete tank with a capacity of about 130 L/SPP. Some of the farms partitioned the tank into two or three interconnected compartments to prevent short-circuiting of flow and to increase the residence time for solids. The cesspits were to be desludged periodically. They were to be a short-term solution until proper waste-treatment facilities were installed. The use of a cesspit was a standard practice imposed by most governments in the tropics.

Seven cesspits, ranging in size from 90 to 270 L/SSP, were investigated to determine their effectiveness in changing wastewater characteristics. Typical results for analyses of data for the first 8 weeks of sampling are given in Table 5.3.

Comparing cesspits using only concentration levels is not inappropriate because the amount of wash water used at the time of the samplings varied from farm to farm. Therefore, the percentage of influent leaving in the effluent was tabulated for each cesspit. The percentage of influent in the effluent for chemical oxygen demand (COD) ranged from 18 to 98%; for TTS, 25 to 137%; for TVS, 16 to 128%. Sometimes the concentration of a parameter in the effluent was greater than in the influent because of turbulence in the cesspit. The ratio TVS/TTS, which indicates the state of solids mineralization, ranged from 0.44 to 0.75.

Table 5.2. Comparison of Singapore waste and wastewater parameters with other sources.

Parameter a

Singapore

(mean)

Malaysia b

Mean

80%

ASAE 1983

(mean)

Taiganides 1977

(mean)

TWF

20

30

-

-

-

TTS

840

90

810

550

690

TVS

670

540

630

440

570

TSS

690

560

660

-

-

COD

840

660

790

510

710

BOD5

250

270

320

180

220

TKN

50

32

41

41

39

a See list of Acronyms and Abbreviations for definitions Units: TWF, L/SPP per day; all others, g/APU per day.

b Source: Teoh et al. (1988). The 80% level was equalled or exceeded 20% of the time.

Using data on raw waste and the results from sampling point 20 (to which all the cesspits under observation drained), it was calculated that cesspits could reduce, from raw pig waste values, COD by 45%, TTS by 42%, and TVS by 33% (see Table 5.1). TTS in the sludge of the cesspits ranged from 3 800 to 164 000 mg/L with an average of 64 000 mg/L. The average ratio of TVS/TTS for sludge was 0.54 (0.20-0.77); for the mat of floating aerosols and solids, which invariably formed in the cesspit, the ratio was 0.73 (0.64-0.83). This indicates that the floating solids in the mat were not mineralized to the same degree as the solids in the bottom of the cesspit.

The large commercial farms in Phase I PFA were required to have pits under the slatted floors. These pits provided a longer storage period than the cesspits in the small farms. However, the effect on waste properties was about the same. BOD5 was reduced by more than 50% in the pit, and TVS was reduced by 32 to 58%. The characteristics of the wastes in the pits under the slatted floors are given in Table 5.4.

Effect of Sedimentation

At the experimental pig production facilities at PPRTI, sampling was carried out almost daily on the wastes that were used in the various treatment experiments. Wastewater from the holding tanks was pumped daily through four parallel fiberglass sedimentation tanks. The characteristics of the wastewaters and the supernatant after settling are summarized in Table 5.5. The daily BOD generation rate was estimated as 0.13 kg BOD/APU. This value is similar to that calculated from the monitoring of wastewater discharge from commercial farms in the Ponggol PFA.


Table 5.3. Cesspit performance sampling data.

Table 5.4. Characteristics of wastewaters from pits under slatted floors.

 

Hold and discharge

a(sluice gates) b

Overflow a

(weir) b

Volume of wastewater (L/APU per day)

46.7

47.0

BOD (kg/APO per day)

0.08

0.10

TVS (kg/APO per day)

0.23

0.37

a System to manage excess.

b Flow discharge.

Table 5.5. Characteristics of wastewater at PPRTI.

Parameter a

Fresh

(mg/L)

Supernatant

(mg/L)

Reduction

%

TTS

5633 ± 205

2809 ± 94

50

TSS

4245 ± 156

733 ± 68

83

BOD5

2665 ± 128

1514 ± 88

43

COD

6255 ± 272

2049 ± 108

67

TKN

-

256 ± 12

-

TAN

-

168 ± 8

-

Note: Values represent the average (+ standard error) of over 200 observations. 'See list of Acronyms and Abbreviations for definitions.

PFA Wastewatershed Characteristics

Pig farming areas as they were planned in Singapore in the early 1970s and as they emerged in Malaysia and other countries were clusters of farms along small creeks in small watersheds that would become "wastewatersheds" (WWSs). The two main advantages of these sites when treated as WWSs were the capability of the natural canal to carry away the farm wastewaters and the access to upstream runoff water. In such sites, PFAs become WWSs that can adopt central treatment plants. The main advantage of centralized treatment is economy of scale. Farmers a]so like the idea because they need not get involved in the operation of processes with which they have no practical experience. A major disadvantage is that the treated effluent cannot be recycled on the farms because it would spread diseases. Reuse of treated water for waste flushing is recommended only on closed-herd farms.

One PFA lent itself to a comparison of centralized and on-farm treatment. The Ponggol PFA and its pollution survey network, the WWS area for sampling point 19, and a typical farmlot layout are shown in Fig. 5.1. Extensive sampling was carried out at sampling point 20 because of the physical shape of the drain at that point; the data were extrapolated to point 19, which had a larger WWS area and was in a low region where a treatment plant could be built. The data collected were used to design, in sufficient detail, 19 treatment combinations and to determine the cost of each. The approaches taken and some of the data collected are highlighted because of their relevance to similar situations elsewhere.

Wastewatershed 19

There were 86 farm lots draining into the WWS at sampling point 19 (WOOS 19), but only 76 lots were actually occupied. Each lot was a full closed-herd farm that operated independently from its neighbours and was enclosed by a fence. Each farm had a cesspit. Effluent from the cesspits flowed by gravity in concrete drains through sampling point 20, past point 19, and into the mangrove area of the Serangoon River. The logical place to build a pilot treatment plant was at sampling point 19.

An engineering survey of the WWS included the tracing of drains, detailed topographical mapping, determining animal population, recording type of production, calculating the roofed area, and measuring drainage area. The total drainage area was 62 ha, of which 19 ha contained roofed barns and buildings. Roofed area and water storage area amounted to 31% of the total land area of the watershed. Assuming 100% collection of precipitation and no losses, rains would yield 484 500 m³ of water annually, estimated to represent 75% of annual demand. The rest of the water had to be supplied.

Pig Population

Until the outbreak of Aujesky's disease in March 1978, actual SPP was in agreement with projected SPP. During March and April 1978, SPP was declining at a rate of 250 SPP/day, mainly from the death of piglets. Consequently, the average live weight per SPP increased during that period; the live weight per pig returned to the normal average of 54 kg once the disease was brought under control.

Wastewatershed Sampling

Weekly sampling was initiated in May 1976 and carried on until May 1978. During this period, in situ measurements and sample collections were made three times a day between 1000 and 1500. Three grab samples of equal volume were collected and composited.

Three series of 24-h continuous samplings were carried out for periods of 5, 4, and 3 days in January, September, and October 1977. In January, the flow was determined by measuring the depth of water and silt in the drain and the velocity of the flow. Grab samples were collected every 30 min and composited. In September and October 1977, however, an automatic sampler was used in conjunction with a 60 V-notch weir. Samples were taken automatically in proportion to the flow, which was automatically and continuously monitored. Because the data collected using the automatic sampler were more accurate than those collected in January 1977, they were used in developing design parameters.

Wastewatershed Design Data

Wastes defecated by the pigs, wash water, base flow from the watershed, and surface runoff water were assumed to constitute the total wastewater flow of WWS 20. The flows at night from 2100 to 0500 were assumed to constitute the base flow. It was also assumed that the base flow was proportional to the watershed area and that it would continue at the same rate during daylight hours. Fortunately, it did not rain during the three 24-h sampling periods, so the wastewater flow could be calculated by subtracting the base flow from the measured flow. The daily volume of wastewater generated from the pig pens was calculated as 36 L/APU or 19.4 L/SPP.

The observed, calculated, and designed values of the daily quantities of the major pollutants at WWS 20 are given in Table 5.6. Almost 58% of the BOD5, 63% of the COD, and 61% of the TTS theoretically defecated by the pigs were either retained in the cesspits or lost, and thus were not found at the sampling point. These reductions and the age of the wastewater need to be considered in the selection of treatment processes.

Flow Pattern

The peak flow of wastewater and pollutants occurred around noon. The peak runoff rate for a 23-min storm (equal to the time of concentration of WWS 19) and 10-year recurrence interval was almost 100 times larger than the peak wastewater flow of 5 m³/min. It is not economical to design a pumping station and a treatment plant catering to such variations in flow. This is the main disadvantage of centralized systems for a WOOS.

By not treating rainfall runoff that exceeded a rate of 5 m³/min, 16.5% of the pollutants would bypass the plant. In other words, 83.5% of the pollutants going through point 19 would be treated during storm runoff periods. Because of the dilution, the water quality of the bypassed wastewater would be acceptable. It was recommended that the open drain be designed with a weir bypass for flows above the peak wastewater flow during periods of no rain.

Table 5.6. Observed mean values at WWS 20 and adopted design values.a

(Apr. 1977 Pollutant parameter b to Oct. 1977)

Observed mean values

WWS Overall mean

Adapted design value

 

Weekly

Daily (24-h)

   
   

Sept.1977

Oct. 1977

   

Dry weather flow

-

-

-

46

46

(base flow + TWF)

         

BOD5

0.102

0.101

0.114

0.106

0.10

COD

0.260

0.357

0.328

0.315

0.32

TTS

0.328

0.331

0.330

0.330

0.33

TVS

0.190

0.209

0.205

0.201

0.20

TSS

0.219

-

-

-

0.22

SLS

3.02

-

-

-

-

TKN

-

-

-

0.039

0.04

TAN

-

-

-

0.029

0.03

a The adopted design values are for the WWS waste flow that goes through the cesspits first; as such, they are lower than the quantities of raw wastes generated on the pig farms. It should be noted that the daily base flow was lo L/APU; thus, the total wastewater flow (TWF) would be 46-10 = 36 L/APU per day.

b Units dry weather flow L/APU per day; SLS, mL/L; all others, kg/APU per day. see list of Acronyms and Abbreviations for definitions.