Municipal solid waste management as a network: what a University of Pannonia model reveals about Hungary’s 2035 targets

Municipal solid waste management in Hungary is not a matter of individual sites but of a national network: every day, waste from 3,154 settlements sets off towards transfer stations, sorting plants, mechanical-biological treatment facilities, composting sites, incinerators and landfills. Researchers at the University of Pannonia have rebuilt this network as a dynamic multilayer model and run the government’s planned measures through it: separate collection of the organic fraction, the mandatory deposit return system, and a new 300,000-tonne incinerator.

The study, published open access in Cleaner Logistics and Supply Chain in July 2026, delivers a sobering result: even combined, the planned steps cut emissions only moderately, and according to the authors they do not guarantee that the 2035 EU targets will be met. Here is how the model works, what the numbers show, and where they deserve a critical reading.

In short: A study modelling Hungary’s municipal solid waste management system as a network finds that separating the organic fraction of mixed waste would cut carbon dioxide emissions the most, while a new incinerator would worsen the balance. Together, the three measures would bring the sector’s footprint below its 2021 level to about 2.02 million tonnes by 2035, but the authors conclude this is still not enough for the EU targets.

Who wrote the study, and why does it matter?

The paper is authored by researchers from three units of the University of Pannonia in Veszprém: Attila I. Katona, Zsolt Szigetvári, Csaba Hegedűs, Zsolt T. Kosztyán, Róbert Kurdi and Viktor Sebestyén (corresponding author). It appeared in volume 20 of Elsevier’s Cleaner Logistics and Supply Chain (article 100354) under a CC BY 4.0 licence, so anyone can read it freely. The research was funded by Hungary’s Ministry of Culture and Innovation through the National Research, Development and Innovation Fund, with two authors also supported by a Bolyai Research Scholarship and an OTKA grant.

The novelty lies less in the carbon balance itself than in the approach. Reviewing the scenario-analysis literature, the authors found that no earlier study handled all four of the following at once:

• national, regional and settlement levels within a single model;
• separately collected waste fractions;
• the system as a genuine transport network with real road distances;
• forecasts of future waste volumes.

The Pannonia model covers all four, and it is built on Hungarian data and Hungarian policy measures.

What is a dynamic multilayer network?

In network science, the elements of a system are nodes and the links between them are edges. In this model the nodes are the actors of the waste system, each with geographic coordinates:

• waste generation points (settlements): 3,154
• transfer stations: 45
• mechanical-biological treatment (MBT) facilities: 23
• selective sorting plants: 56
• composting plants: 60
• biogas plants: 77
• landfills: 37
• incinerators: 11

The edges are directed: they show where waste goes, and how much of it, in a given year. “Multilayer” means that different types of material flow sit in separate layers. The model distinguishes ten, such as settlement–transfer station, transfer station–MBT, settlement–sorting plant, MBT–landfill, MBT–incinerator and settlement–biogas plant. One physical facility can appear in several layers: an MBT, for instance, receives mixed waste from settlements and residues from sorting plants, then sends outputs on to landfill and incineration.

“Dynamic” refers to time: the network starts from the 2021 base year and is updated year by year up to 2035. A separate network instance is built for each scenario and each year, so changes in flows and emissions can be traced for every intervention.

Where does the data come from, and how does the model calculate?

The network draws on three main sources: settlement-level waste data from the Hungarian Central Statistical Office, facility data from the National Environmental Information System (OKIR), and tables provided by MOHU, the national waste management concession company. Actors’ addresses were geocoded with a Python script, and road distances were calculated with OSMnx, an OpenStreetMap-based package.

One key methodological choice is that every source is assigned to the nearest suitable facility by road. This is both a strength and a limitation: it shows the geographic optimum clearly, but not necessarily where waste actually goes under real contracts and capacity limits.

Forecasting: why a logarithmic curve?

Missing settlement data were filled by interpolation, and future volumes were projected layer by layer to 2035 using a logarithmic function. The authors tested several methods:

• a linear model produced unrealistically steep growth because it overreacts to recent years;
• an ARIMA model suggested average annual growth of around 2.5%, since it ignores the dampening effect of EU directives;
• the logarithmic model projected roughly 0.52% growth a year and fitted better;
• a deep-learning LSTM network was rejected because of its data hunger and black-box nature.

Emission factors

Emissions come from two sources. For transport, the model uses the Sphera (GaBi) factor for a 20-tonne truck with a 17.3-tonne payload, including the empty return trip. Treatment technologies carry the following factors, in kilograms of CO₂ equivalent per kilogram of waste:

• landfill: 0.838
• incineration: 1.26
• composting (open windrow): 0.508
• biogas and biomethane production: −0.0662, i.e. a net emission reduction

The authors deliberately left out the life-cycle effects of recycling, which would have required separate models for each material and technology. We come back to this in the assessment below.

Seven regions, very different infrastructure

Hungary’s system is coordinated through waste management regions, and the model reveals marked differences between them:

• Észak-Dunántúl (North Transdanubia): 764 settlements, 1.63 million people; 12 transfer stations, 5 MBTs, 14 sorting plants, 6 composting plants, 9 landfills, 1 incinerator.
• Dél-Dunántúl (South Transdanubia): 948 settlements, 1.35 million people; 10 transfer stations, 6 MBTs, 13 sorting plants, 14 composting plants, 6 landfills, 4 incinerators.
• Budapest area: 71 settlements, 2.35 million people; 5 transfer stations, 1 MBT, 3 sorting plants, 3 composting plants, 2 landfills, 3 incinerators.
• Közép-Kelet Magyarország (Central-Eastern Hungary): 243 settlements, 1.40 million people; 5 transfer stations, 3 MBTs, 7 sorting plants, 9 composting plants, 6 landfills, no incinerator.
• Észak-Kelet Magyarország (Northeastern Hungary): 730 settlements, 1.73 million people; 8 transfer stations, 7 MBTs, 14 sorting plants, 15 composting plants, 10 landfills, 1 incinerator.
• Tiszántúl (Trans-Tisza): 398 settlements, 1.42 million people; 5 transfer stations, 1 MBT, 5 sorting plants, 13 composting plants, 4 landfills, 2 incinerators.

Facilities are more evenly spread in the western regions, while in the east a few intermediate plants serve many settlements. The map also mirrors Hungary’s star-shaped, Budapest-centred transport network.

Transport routes: 116 km on average, up to 307 km

According to the model, a waste fraction travels 116 kilometres on average from where it is generated to its final treatment. The longest journeys are all in Tiszántúl: waste from Tiszagyenda (307 km), Örményes (299 km), Kunhegyes (298 km), Túrkeve and Tiszaszentimre (294 km each) travels furthest. The reason is the scarcity of landfills and incinerators in the area, which forces waste into other regions.

Some telling results of the regional analysis:

• around 80% of settlements in Tiszántúl could reach MBTs in neighbouring Northeastern Hungary by a shorter route than those in their own region;
• roughly half of sorting plants are closer to an MBT in another region than to one in their own;
• in the MBT–incinerator layer, almost the entire country shows cross-regional flows because incineration capacity is scarce;
• if every actor shipped to the nearest suitable facility regardless of regional boundaries, total network edge length would fall from 411,233 km to 389,865 km, a saving of just over 5%.

The authors also ran a capacity stress test: if every source had to use its second-nearest facility, both average and total path length would rise about 1.7-fold. Their conclusion is that capacity outages are best handled with regional backup capacity rather than by restructuring the network.

The carbon footprint of municipal solid waste management: transport is only one percent

One of the study’s most important messages, often misunderstood in public debate, concerns proportions. In 2021, total emissions from the Hungarian system were 2,273,293 tonnes of CO₂ equivalent. Of this:

• landfill: 1,344,223 tonnes (59.1%)
• incineration: 747,432 tonnes (32.9%)
• composting: 160,826 tonnes
• transport: 20,812 tonnes (0.9%)

About 77% of transport emissions relate to collection within settlements and haulage to MBTs. MBT–landfill edges account for only 0.12% of the transport footprint, because MBTs were historically built next to landfills.

The conclusion is clear: the bulk of emissions arise not on the road but at the treatment stage. Optimising transport is useful, but what really matters is where the waste ends up. At the same time, the authors note that moving away from landfill may require new facilities and longer routes, so siting decisions need a network perspective.

Five scenarios for 2035

The model evaluates every scenario for 2035, the year to which the key EU deadlines are tied: at least 65% municipal waste recycling and no more than 10% landfill. For packaging waste the target is 70% recycling by 2030, and for single-use plastic bottles 90% separate collection by 2029.

1. Baseline: the system’s 2035 state with no new interventions, extending historical trends.
2. Organic fraction separation (OFMSW): the organic fraction of mixed waste, put at 20% by mass based on laboratory waste analyses, is collected separately and treated in biogas plants. In the model this means 543,403 tonnes of organic waste.
3. Deposit return system (DRS): 158,000 tonnes of useful material are diverted from mixed and selective streams. The study assumes the HUF 50 deposit and 2,657 return points.
4. Waste-to-energy (WTE): a new incinerator with an annual capacity of 300,000 tonnes, sited according to the national distribution of combustible waste. Existing incineration would rise on its own from 388,579 tonnes in 2021 to 410,140 tonnes in 2035.
5. Waste Transition: all three measures combined.

What do the results show? Organics win, incineration loses

Emission results by scenario (2035, tonnes of CO₂ equivalent):

• baseline: 2,433,481 (+160,188 versus 2021)
• DRS: 2,281,425 (−152,056 versus baseline)
• organic fraction separation: 1,916,768 (−516,713)
• new incinerator: 2,629,333 (+195,852)
• combined Waste Transition: 2,024,606 (−408,875)

Separating organic waste is thus the most effective measure: diverting methane-producing organic matter from landfill to biogas plants saves more than half a million tonnes of CO₂ equivalent on its own. The DRS has a moderate effect but yields a clean, valuable material stream. In the model, the new incinerator increases emissions, because incineration’s factor of 1.26 is higher than landfill’s 0.838.

This leads to perhaps the study’s most interesting finding: the combined transition performs worse than organic separation alone. The incineration component “eats back” part of the savings. The authors frame this not as a rejection of incineration but as a trade-off: incineration reduces landfill and transported volumes and diversifies the energy mix, but it is costly in carbon terms.

A note on interpretation: the study’s waterfall chart chains scenario totals one after another, so it shows −347,908 tonnes at the DRS step and −364,657 tonnes at the organics step. These are not the standalone effects of each measure, but differences from the previous bar. The standalone effects are those in the list above.

Transported volumes: decisions at source ripple through the chain

In 2021, the total volume moved along network edges was 6.77 million tonnes. This exceeds the amount of waste generated, because in two-stage collection the same tonne travels several times (settlement to transfer station, then to an MBT, then to landfill or incineration). Under the baseline it would grow to 7.12 million tonnes, but under the combined transition it would fall to 5.90 million tonnes.

The authors call this network simplification: when waste takes a different route right where it is generated, the effect ripples through the whole chain. Transport emissions under the combined transition would be 19,995 tonnes, lower than in 2021 even though waste volumes grow.

Other environmental indicators and costs

Beyond the carbon footprint, the authors compared further life-cycle indicators, with 2021 set at 100. The combined transition would cut global warming potential to 89 and freshwater eutrophication to 72, and keep most other categories roughly level. The exception is water use, which would rise to 158, i.e. about one and a half times, mainly because of organic waste treatment.

The study also includes a simple cost model: transport at €0.0335 per tonne-km, MBT sorting at €33.48/t, incineration at €125.56/t and landfill at €66.96/t, with 3% inflation and an 8% discount rate. Annual system costs would rise from €259 million in 2021 to €417 million in 2035, with a present value of €2,851 million for 2021–2035.

Where will the treatment hotspots be in 2035?

Under the combined transition scenario, the map of final treatment looks like this:

• Biogas plants: this is where the biggest structural change would occur. Seventy-two facilities would come into play, handling over 543,000 tonnes in total; 18% of the volume would fall around Budapest. According to the authors, the spread of plants means no further large hubs are needed.
• Composting plants: close to 400,000 tonnes, 26% more than in 2021; the largest load, around 70,000 tonnes, would again be next to the capital.
• Incinerators: a highly concentrated picture. The two largest, at around 460,000 and 200,000 tonnes (the latter being the new plant), both sit near Budapest and together receive 80% of incinerated waste. The authors suggest future expansion should target western and south-eastern Hungary.
• Landfills: landfilled waste would fall to 31% of the 2021 level, or 501,000 tonnes, with 21 of 37 possible sites remaining in use. The heaviest load, 92,000 tonnes, would go to Hejőpapi, the main collection point for north-eastern Hungary, followed by a landfill near the capital with 85,000 tonnes.

Hejőpapi’s prominence will not surprise anyone who knows the region: alongside the landfill, it hosts mechanical-biological treatment and refuse-derived fuel production capacity, so the model reflects the weight of existing infrastructure.

Are the MBTs in the wrong place?

An appendix to the study presents a separate optimisation test. The authors ran k-medoid clustering on the road distance matrix between settlements with exactly 23 clusters, the same as the number of operating MBTs. This isolates the question of location rather than the number of facilities.

The result: in the western and central regions several MBTs sit close to their cluster’s “ideal” centre, but in the east the optimal points fall far from existing sites. In the authors’ words, the long eastern routes point to a placement problem, not a capacity-count problem. If new capacity is built or investment priorities are set, shifting towards the eastern medoids would deliver the largest transport savings.

What does the study tell policymakers?

The authors make five recommendations:

• Tighten decarbonisation policy: current measures, even if fully implemented, are not enough for 2035 and, in the authors’ view, not certainly enough for 2040 either.
• Integrated, network-based planning: the side effects of individual interventions only become visible when the whole system is modelled, so CO₂ reduction and future capacity needs must be planned together.
• Address regional imbalances: infrastructure gaps, particularly in the east, cause long transport routes.
• Encourage early-stage diversion: separation at source has a multiplied effect at every level of the chain.
• Multi-criteria decision-making: besides the carbon footprint, energy security, resource recovery and social acceptance must also be weighed.

The timing of these recommendations is no accident. Hungary’s institutional framework for waste is being reshaped right now: an amendment to the waste act is in public consultation, and a new top-level environmental authority will oversee waste management from 2027. A national, facility-level network model could provide an objective benchmark for such decisions.

A professional view: strengths and question marks

What is convincing

The study’s greatest merit is that, for the first time, it depicts the Hungarian system as a single network broken down to settlement level and built on real road distances. Placing transport and technology emissions side by side draws attention to what matters: the climate impact depends overwhelmingly on the choice of treatment route. Quantifying the gaps in the east, the 80% figure for Tiszántúl and the k-medoid analysis offer a concrete, mappable basis for debate.

It is also notable that the model’s assumptions can be checked against reality. The study assumes 158,000 tonnes diverted by the DRS; according to MOHU data, 67,952 tonnes of beverage packaging were returned in the first half of 2026, or roughly 136,000 tonnes on an annual basis. The order of magnitude matches, and as the system grows the assumption looks realistic. (The number of return points, meanwhile, is well above the study’s 2,657: by June 2026 there were 5,065.)

What deserves a critical reading

The incineration emission factor. The 1.26 kg CO₂e/kg value comes from a generic European dataset. The study does not specify whether it includes credits for displaced electricity and district heat. If not, incineration’s relative disadvantage in the model is larger than it would be in a full-system life-cycle assessment. From the other direction, real costs will also hit incineration in future: from 2031, the EU emissions trading system is expected to cover waste incineration, which the cost model does not yet reflect.

The fate of RDF. In Hungarian practice, a large share of the refuse-derived fuel produced at MBTs goes to co-processing in cement kilns, where it replaces fossil fuel. The model applies the incineration factor uniformly to MBT–incinerator flows; back-calculating from 2021 incinerator emissions suggests the “incinerator” category covers RDF as well as direct incineration. Accounting for the substitution effect of co-processing could give a more nuanced view of the WTE scenario.

Recycling left out. Because the life-cycle effects of material recycling are not modelled, the climate benefit of the DRS and selective collection is underestimated. The authors acknowledge this themselves.

Internal inconsistencies. Several figures do not match within the paper. The 2035 baseline total transported volume is given once as 7.12 and once as 7.4 million tonnes. The text says landfilled volume falls in the baseline “due to DRS”, although the baseline specifically excludes the DRS. The 543,403 tonnes of organic waste are 20% of 2021 mixed waste, not of the 2.17 million tonnes projected for 2035 (which would give about 435,000 tonnes). For Tiszántúl, the facility list includes two incinerators, while the text points to a lack of them. The limitations section cites the absence of cost modelling, even though the paper presents a net present value calculation, which is in fact the present value of system costs.

The landfill target. This is the most important point. According to the map-based analysis, 501,000 tonnes would be landfilled under the combined transition. Yet back-calculating from landfill emissions in the carbon table (818,331 tonnes) with the 0.838 factor gives roughly 0.98 million tonnes. By our own estimate, against municipal waste of some 3.5–3.8 million tonnes a year, the lower figure implies a landfill rate of about 13–14% and the higher one around 25%. Whichever is closer to reality, the EU’s 10% ceiling is not met in the model, which reinforces the authors’ main conclusion. We explained why landfill is hard to curb where it is cheap in our European comparison of landfill taxes.

The nearest-facility principle. The assignment shows a geographic optimum, not actual flows shaped by contracts and capacity limits. The authors treat this as a limitation and leave capacity planning to future research.

The wider context

The model’s results fit the broader picture painted by national statistics. According to the latest Eurostat data, Hungary’s recycling rate excluding mineral waste is just 33.6%, and its landfill rate is well above the EU average, meaning circularity of municipal streams lags far behind its potential. On the organics side, the technology already works: the Szarvas biomethane plant has shown that biogas can be injected into the grid.

But the study’s 72-biogas-plant scenario also requires clean, separate collection of household organic waste, where Hungary so far has only pilot experience. The Czech Republic, whose 2035 pathways the study cites as a reference, faces a similar challenge, with 40% of its mixed municipal waste being biodegradable. Cross-boundary flows are not a purely Hungarian phenomenon either: our analysis of cross-border waste shipments shows how flow directions can distort the statistical picture.

Conclusion

The University of Pannonia study gives Hungarian waste planning a new tool: a network model that tracks waste from settlement to facility and quantifies emissions scenario by scenario. Its main messages are clear. Most of the climate impact arises at treatment, not in transport. Separating organic waste is the most effective step, while new incineration capacity worsens the carbon balance. Infrastructure in eastern Hungary fits poorly with the geography of waste generation. And the planned measures, even combined, do not secure the 2035 targets.

Given some of the model’s assumptions and internal inconsistencies, the exact figures should be treated with care, but the directions are supported by national data. Perhaps the most important lesson is that interventions must be designed as a system rather than one by one, and for that there is now an open-access methodology built on Hungarian data.

Frequently asked questions about the municipal solid waste management network model

What did the University of Pannonia study examine?

The study modelled Hungary’s municipal solid waste management system as a dynamic multilayer network. It linked 3,154 settlements and all treatment facilities through real road distances, forecast waste volumes to 2035, and calculated carbon dioxide emissions under five scenarios: baseline, organic fraction separation, deposit return system, a new incinerator, and all three measures combined.

Which measure cuts waste management emissions the most?

Separating the organic fraction of mixed waste and treating it in biogas plants. According to the model, this would reduce 2035 emissions by 516,713 tonnes of CO₂ equivalent compared with the baseline, because biomethane replaces methane-producing organic matter in landfills. The deposit return system would save 152,056 tonnes, while a new incinerator would add 195,852 tonnes.

Why does the combined scenario perform worse than organic separation alone?

Because the combined scenario also includes the new incinerator. In the model, incineration’s emission factor (1.26 kg CO₂e/kg) is higher than landfill’s (0.838), so the incineration component eats back part of the savings. The combined transition would bring 2035 emissions to 2,024,606 tonnes, whereas organic separation alone would reach 1,916,768 tonnes.

How large is transport’s share of the waste sector’s carbon footprint?

Very small: in 2021, transport accounted for only 0.9% of the system’s 2,273,293 tonnes of emissions, or 20,812 tonnes. Landfill represented 59.1% and incineration 32.9%. About 77% of transport emissions came from collection within settlements and haulage to MBTs, so the climate impact depends mainly on the treatment route chosen.

Why does waste travel furthest in the Tiszántúl region?

Because the area has few landfills, incinerators and mechanical-biological treatment plants, so waste must be shipped to other regions. The longest journeys start in Tiszagyenda (307 km), Örményes, Kunhegyes, Túrkeve and Tiszaszentimre. The model shows that around 80% of Tiszántúl settlements could reach MBTs in Northeastern Hungary by a shorter route.

Will Hungary meet the 2035 EU landfill target according to the model?

According to the model, no. Under the combined transition, the map-based analysis gives 501,000 tonnes landfilled, while back-calculating from the emissions table gives roughly 0.98 million tonnes. Against 3.5–3.8 million tonnes of municipal waste a year, both values exceed the EU’s 10% ceiling, which reinforces the authors’ main conclusion.

🔗 Official sources and references: 👉

• Katona, A. I., Szigetvári, Z., Hegedűs, C., Kosztyán, Z. T., Kurdi, R., Sebestyén, V. (2026): Evaluation of municipal solid waste management measures in Hungary: A dynamic multilayer network framework. Cleaner Logistics and Supply Chain, 20, 100354. DOI: 10.1016/j.clscn.2026.100354 (open access, CC BY 4.0)
• Hungarian Ministry of Innovation and Technology (2021): National Waste Management Plan 2021–2027 (in Hungarian)

Consistent structure, quick answers: every article is built so that both readers and AI search tools can find the key points easily. Our content is produced with the help of AI tools and reviewed by our editor.