Summary and conclusions, Miljøstyrelsen

Nutrient Removal in Small Treatment Plant with Urine Separation in the Catchment Area

Summary and conclusions

This investigation of the effect of nutrient removal in a small treatment plant with urine separation in the catchment area was carried out at Svanholm, an estate housing a cooperative with about 110 residents. It is a project under the action plan on sustainable urban renewal and wastewater treatment, Theme 3 – Assessment of the potential for and limitations to recycling of nutrients from urban to rural areas.

Separate collection of urine for fertilisation purposes reduces the content of nutrients in the remaining sewage. The investigation examines the extent of the reduction and its impact on the operation of treatment plants with requirements for nutrient removal.

The project was carried out at the estate of Svanholm. The residents of the estate operate their own treatment plant with nutrient removal, and project on exploitation of nutrient from urine was previously completed at the estate. Therefore, the residents of the estate have been in a favourable position to conduct the practical part of the project, which follows up on the previous activities.

The project included the entire sewage system at the estate of Svanholm.

Two different types of urine separating toilets have been established in connection with the common dining hall of the estate. They represent the systems that were available on the market in Denmark at the start of the project. The residents of the estate have evaluated the systems and concluded that both types are usable; however, improvements are called for, if they are to work optimally.

The sewer system was mapped jointly by the residents of the estate and Erling Holm aps, and a detailed outline of sewers and toilets in the entire catchment area of the treatment plant has been drawn up. The mapping showed that 32 toilets are located in seven buildings/groups of buildings and that two toilets are located at some distance from the others.

In connection with the project, continuous water flow measurements were made at the treatment plant. Thus, water flow variations were mapped, and it appeared that the sewer system generally works as a separate sewer system, showing only minor water flow increases in connection with precipitation.

A draft of a system for complete urine separation in the catchment area shows that a suitable solution would be to divide toilets into four groups, each with a separate collection system, and to establish separate treatment of the entire discharge from the two remote toilets, disconnecting them from the sewage system.

The treatment plant of the estate is a traditional recirculation plant for nitrogen removal and simultaneous precipitation of phosphorous. Residents of the estate built it in 1993. In addition to the residents of the estate it serves a number of small manufacturing enterprises and the common kitchen of the estate. Design and operation of the plant have been mapped for the purpose of computer simulation of operation under different assumptions as to the degree of nutrient reduction from urine separation in the catchment area. An initial load survey of the treatment plant compared with historical load measurements shows that occasionally, the plant received sewage with a very high organic matter content. Mapping of enterprises and institutions in the catchment area revealed that the reason might be spillage to the sewer system from the milking area of the estate. A minor alteration of the system ensured that this spillage has been stopped. A subsequent load measurement shows sewage concentrations corresponding to normal domestic sewage.

Based on loads on the treatment plant and empirical data for contributions from urine and other sewage flows, three scenarios have been drawn up for the future load on the plant after establishment of urine separation in the catchment area.

The baseline scenario is the present situation without establishment of urine separation.

Two scenarios describe the option of urine separation in the entire catchment area. If the system works at 100 per cent efficiency, the load of nitrogen is reduced by 80 per cent and that of phosphorous by 50 per cent. In a more realistic assessment of the potential for separate collection of urine, a two-thirds reduction of nitrogen and a 40 per cent reduction of phosphorous are anticipated. Water and organic matter contents in urine are very low, and consequently the organic matter load and the hydraulic load are not changed substantially after the establishment of urine separation.

The three scenarios were computer-simulated with the simulation programme EFOR 2001. After calibration of the model corresponding to the present load and present process conditions, the model was used to simulate the impact on the treatment plant from a reduction of the load resulting from the above scenarios.

In addition to simulations of the immediate effect of a reduction of the load, calculations have been made of possible savings of electricity and chemicals from optimisation of plant operation under reduced load conditions. Furthermore, calculations have been made elucidating the possibility of providing sufficient treatment capacity to comply with current discharge requirements by major reductions in plant volumes, since in particular the lower nitrogen load reduces the need for capacity for oxidation of nitrogen compounds.

Computer simulations show that with a 100 per cent efficient urine separation system amounts of nitrogen are insufficient to maintain the biological processes in the treatment plant. In practice it is not possible to ensure such an extensive separation, so problems would hardly arise in practice. However, calculations show that in the separation of sewage a balance must be secured between nitrogen and organic matter, if extensive biological reduction of organic matter is wanted.

In a realistic assessment of an efficient urine separating system corresponding to a 80 per cent separation of urine, the load on the treatment plant will be reduced substantially, and the treatment for both nitrogen and phosphorous will be improved significantly without altering the operation of the plant.

Thus, urine separation in the catchment area of a treatment plant must, also to a minor extent, be expected generally to result in improved treatment.

Optimisation of chemical precipitation can lead to considerable savings in amounts of chemicals. Considerable phosphorous amounts are bound in sludge regardless of urine separation, and consequently savings in chemicals will be larger than the relative reduction in the phosphorous load.

Optimisation of operation can be achieved in several ways. There are immediate savings in electricity from the reduced ammonium load. The internal recirculation that is part of the denitrification process can be discontinued, as nitrate amounts to be denitrified are limited. Sludge age in the plant can be reduced to a level where no nitrification takes place, so that electricity consumption for oxidation of ammonium can be avoided completely.

Computer simulations show that only a limited increase in discharges of Total-N as nitrate will occur if the internal recirculation is closed. If nitrification is stopped, discharges of Total-N will increase modestly, though in this case by larger discharges of ammonium. Neither of the solutions will cause problems in connection with complying with discharge requirements for Total-N for the treatment plant at the estate of Svanholm; however, a complete nitrification stop brings ammonium discharges close to plant discharge requirements, so that it is doubtful whether compliance will be achieved in practice.

The plant will comply with current discharge requirements with a considerable margin after establishment of a comprehensive urine separation system, and consequently the capacity of the existing plant will be considerably larger than needed. Therefore, computer simulations have been made of the consequences of reducing volumes of the plant. As the need for denitrification is limited, the denitrification tank and the internal recirculation of the plant can be removed. As the need for nitrification is limited, the aeration tank volume can be reduced substantially.

Simulations show that the denitrification tank of the plant and the internal recirculation can be removed without significant increases of Total-N discharges. Thus, the plant can comply with discharge requirements without any problems. A reduction of the aerated volume to one third of the present volume will reduce sludge age to a level where nitrification stops. The effect corresponds roughly to the effect from reducing sludge age as explained above, so that also here it is doubtful whether the plant can comply with current requirements for ammonium.

Based on the draft for establishment of urine separation at the estate of Svanholm and the computer simulations, the costs and savings potentials can be calculated.

Total costs for establishment of urine separation at the estate of Svanholm have been estimated at around DKK 1.5 million. Half of this amount covers expenses for toilets, collection tanks and other equipment, the remainder being costs for contractors. The latter item in particular is difficult to estimate, as new installations in old buildings can be very difficult and costly.

The potential for savings in the operation at the present plant is limited. It has been estimated that an immediate saving in chemicals and electricity of around DKK 5,000, corresponding to 20 per cent of present operating costs, can be expected from the establishment of urine separation.

If the plant is operated without nitrification, savings are slightly higher, but the ability of the plant to comply with discharge requirements becomes less certain.

With new construction of a treatment plant it is possible to save considerable plant volumes through the establishment of urine separation in the catchment area, and it will probably be possible to secure very low discharges of both organic matter and nutrients with more simple treatment techniques than those applied at the estate of Svanholm today.

Tax reductions relating to the discharge of organic matter, nitrogen and phosphorous from the establishment of urine separation at the estate of Svanholm will only be limited, and they can hardly be realised simultaneously with an optimisation of operating costs.

With the proposed system for urine separation at the estate of Svanholm an estimated 370 kg of nitrogen and 50 kg of phosphorous will be collected annually. Urine has a high fertilisation value; however, in view of the costs of transportation and application, it will probably not contribute to improving the system economy.

However, the costs of establishing urine separation at the estate of Svanholm will be high and cannot by offset by savings achieved in the operation of the present treatment plant and income from sale of urine as fertiliser.

The financial potential will be substantially higher when such systems are established in new building projects, where the costs of a urine separation system will be lower, and where the significant reduction of contents of nitrogen and phosphorous in sewage can be used to build a more simpler and less costly treatment plant.