Sections

Services

Information

UK United Kingdom

The carbon devil in the detail on urban density

How dense could we be? Very, if you follow much of the commentary in Australian debates about the way we should plan our cities. High-rise residential developments have been springing up in all Australia’s…

Is Australia going down the East Asian high-rise route? eugene

How dense could we be? Very, if you follow much of the commentary in Australian debates about the way we should plan our cities.

High-rise residential developments have been springing up in all Australia’s major cities. The view that carbon constraint, ecological protection and liveability can only be achieved by remodeling our cities at high-rise densities has taken root among much of Australia’s policy intelligentsia. This view is inconveniently flawed.

Most people agree climate change should be understood through robust scientific evidence. Such a standard should also apply to measures to reduce the climate impact of our cities.

Unfortunately a simple formula equating high-rise urbanism with low carbon or ecological impact finds at best partial confirmation in the scientific literature.

When different building scales are compared on objective environmental criteria the evidence suggests that high-rise apartments are often the worst performers. The building scale with least overall ecological impact – measured in energy, CO₂ and water use per capita – tends to be medium-rise of between three to six storeys, with individual detached dwellings the next best.

Alan Perkins and colleagues' work in Urban Policy and Research, for example shows that on per-capita analysis, attached low-rise dwellings perform best in terms of CO₂, with suburban and high-rise successively worse.

Such patterns arise because any valid assessment must account for all energy use. This includes energy “embodied” in a building during construction, plus energy used in ongoing operations.

Medium-rise residences tick a lot of low-emissions boxes. pov steve

While residents of detached houses use high levels of energy to get around, because of their greater car reliance, their overall energy use per capita tends to be lower than high-rise residents. This is because detached dwellings can consume less embodied and operational energy and that use is divided among a higher average occupancy rate (detached dwellings generally house more people than apartments).

The relationships between building type, urban form and CO₂ emissions become even more complex when income – or “lifestyle” – factors are included in the analysis. But it seems that high-rise urbanism exacts a high carbon cost.

The Australian carbon consumption atlas, prepared for the Australian Conservation Foundation by Chris Dey of the University of Sydney and colleagues, provides a striking illustration of this pattern.

Their work shows the highest per capita residential environmental consumption occurs in the higher density inner urban areas of Australian cities. The 37 tonnes of total CO₂ consumed per person each year by downtown Sydney residents is, for example, more than double the 16 tonnes produced by residents of Blacktown. There’s a carbon devil in the detail on density.

Such findings fundamentally confound the simple “high density good, low density bad” assumptions in current debates. High-rise apartments are far less of a solution to our urban environmental challenges than the prevailing consensus suggests. Even if better design could improve high-rise performance, so too could it improve that of other building types.

Some look to New York City as a model for Australian cities. But high-rise Manhattan is a small island set among a continuous interconnected urban mega-region of some 21 million inhabitants, only some of whom live at high densities.

Paul Mees of RMIT has compiled and compared official data on the density of North American and Australian cities. His peer-reviewed work shows that the densest city in the United States is not New York but Los Angeles, which few Australian urbanists aspire to emulate.

Sydney, Australia’s densest city with 20 persons per hectare, is only three quarters as dense as Los Angeles, which has 27 persons per hectare. Yet public transport commuting in Sydney is five times the Los Angeles rate. Transport policy, Mees argues convincingly, is at least as important as urban form in shaping a city’s transport outcomes. Density is not destiny.

Density doesn’t determine how well public transport works. xlynx

With high-rise urbanism facing scientific doubts, its advocates often decorate their arguments with circumlocutions about the aesthetic or cultural qualities of high-rise living. These claims are usually subjective and contestable. For every high-rise architectural masterpiece there are dozens of shoddy apartment blocks thrown up with little ecological or aesthetic concern.

The supposed “vibrancy” of high-rise living is often balanced against provision of green open space, especially at the neighbourhood scale. Children’s development advances through free physically active play but a recent literature review by my colleagues Jason Byrne and Neil Sipe argues planning for higher densities rarely considers children’s needs. Blubberville anyone?

Why then have views favouring high densities gained traction? Despite the waning of the old cultural cringe, a view persists of Australia’s essential suburban character – where most of us live - as artless against the boulevards of Europe or the boogie-woogie of New York.

The intellectual misadventure of high-rise urbanism also perpetuates a pernicious bias in Australian environmental debates in which less affluent suburban dwellers are treated as environmentally unsophisticated “bogans” – a stereotype recently denounced by Melbourne University’s David Nichols.

It fits within a long and regrettably continuing Australian tradition of denigrating suburbia whose recent version sneers at “aspirationals” in suburban “McMansions” driving “monster-trucks”. That complaints about suburban consumption lack objective scientific foundation, raises suspicions that the anti-suburban prejudice serves to deflect scrutiny from the more harmful consumption patterns of wealthier – and typically denser – inner urban households.

Those who criticise high-rise urbanism, though, risk being cast as apologists for urban sprawl. Disagreeing with Sydney’s Barangaroo proposal, for example, doesn’t equate to support for the latest fringe growth area splurge.

More single, detached dwellings in low density estates at the suburban fringe also causes harms. These range from the destruction of bio-diverse habitats to the social isolation of new residents from work and services. My own work on household oil vulnerability clearly reveals the future perils from higher fuel prices already planned into the fabric of many of our car-dependent fringe suburban zones.

So how dense should we be? One possibility is a mid-level suburbanism. Focused nodes of three to six storey development – terraces, walk-ups and low-rise apartments – are dispersed across an entire metropolitan area, not just inner zones. Tied to an employment decentralisation program where more jobs exist outside the CBD, such nodes can then be used to anchor improved suburban public transport networks.

More jobs and medium-density housing in suburbs could answer some planning questions. La Citta Vitta

As Paul Mees has shown, suburban densities can easily support high quality public transport if we plan our networks effectively. Such a model can accommodate new residents more widely across the city, and avoids the ecological damage of very high or low densities.

There is an even better solution to the density debate though. Maybe we should abandon our obsession with density as a driving objective of urban policy and instead design our living environments to best accommodate our overall economic, social and environmental goals, at whichever scale is most appropriate.

This means viewing density as just one instrument among many in our urban toolkit. We should use it carefully and selectively, not as a universal prescription that is forced unthinkingly upon our cities.

It similarly requires us to understand the form of our cities as the effect of urban social, economic and institutional processes, not the determinant. People congregate and interact in cities because of social and economic logics, not because they were made to by urban form.

Refreshingly, the draft Australian national urban design guidelines shy away from installing density as a driver of urban outcomes and instead propose methods that could operate at any scale of urban form, albeit rather innocent of the social and economic processes shaping cities and their appearance.

We could be very dense in the way we plan our cities. But we can be much cleverer than that.

Articles also by This Author

Sign in to Favourite

Join the conversation

27 Comments sorted by

  1. Mark Harrigan

    PhD Physicist

    Thanks, interesting article - but I'm having a little trouble reconciling some of the conclusions with publicaly available data

    On pages 45-47 of the UGEC report on cities and climate change (http://www.ugec.org/docs/Cities%20and%20Climate%20Change%20book.pdf) it shows quite clearly that relatively less dense city like Denver has emissions per capita of 19.4 tonnes versus 10.5 for New York city (nearly double) - Even Los Angeles is only 13.

    I do not see how it is possible to reconcile that data…

    Read more
    1. Ray Pritchard

      Graduate Engineer

      In reply to Mark Harrigan

      I think that comparing cities on their carbon footprint can be done to support either side. Los Angeles is a very low density city with rather poor public transport (there's a reason their highways look like this: http://www.worldofstock.com/slides/TRC4904.jpg)

      Having not gone into any great detail beyond what is written above I could see how medium-rise apartments may deliver the best results in this test, but I think that the conclusion based upon Sydney's downtown residents consuming over twice…

      Read more
    2. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Mark Harrigan

      Certainly we can compare cities at aggregate density but there's devils in the details almost every time we do that. Wealth and occupancy rates create considerable diversity in residential consumption patterns at the household level and the evidence seems to indicate that when these are accounted for high-rise development performs poorly in contrast to other scales.

      report
    3. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Mark Harrigan

      @ Mark above (apologies - for some reason the system won't append this reply directly beneath your comment). The data in the UGEC report appears to be concerned with emissions from companies rather than households so differences is the economic base of the cities could exert considerable influence on overall GHG emissions. Mees' data shows Denver is nearly half as dense as LA in population terms but has almost identical transport mode shares - although those don't translate directly into GHG they indicate that factors other than urban form are likely at play in determining emissions.

      report
    4. Mark Harrigan

      PhD Physicist

      In reply to Jago Dodson

      Thanks Jago - just goes to show that this a complex issue and that any analysis needs to avoid oversimplification to avoid generalisations.

      It appears to me, not being an expert by any means but having done some reading on the matter (again I really encourage you and others to get hold of the September Scientific American on this - it's quite informative) that taking a holistuc approach to urban design principles, but also allowing the community to take bottom up (rather than directed top down…

      Read more
  2. Eclipse Now

    Manager of Graphic Design firm

    Individual energy use IN the high rise building is only one point to consider. There are a host of others.

    * New Urbanism is more efficient with materials, more effective at creating local communities, and more efficient with your time. How much time did you lose in traffic today?
    * It's not just an energy efficient house; but an energy efficient city that drastically reduces the number of cars required in the first place.
    * You could live most of the time without a car, and hire…

    Read more
    1. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Eclipse Now

      If new urbanism is more efficient on energy and carbon then I'd expect to see some evidence in the literature. Defining new urbanism to give it scientific purchase would be a good start.

      report
  3. Eclipse Now

    Manager of Graphic Design firm

    Let's not forget that LA has been described as a city looking for a CBD. Sydney public transport at least has somewhere to 'go', to aim for. The CBD and it's partners of Chatswood and Parramatta etc.

    Where would LA take them? What major work corridors exist in LA? I think the LA comparison is poor, but agree with the overall emphasis on public transport. We're going to need it. Peak oil is here, and terminal decline begins anywhere from now to about 5 years away.

    report
    1. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Eclipse Now

      As noted, few urbanists seek to emulate LA, and for a raft of reasons unrelated to its density. But Mees' data shows relying on density as an indicator or determinant of urban transport energy use is problematic.

      report
  4. Geoff Russell

    Computer Programmer, Author

    Regression analysis 101 ... control for confounders. Using the consumption atlas to argue about impacts of housing density is silly without considering variables like, for example, income. People with more money buy more stuff, and even if they live in medium density buildings, they travel more, if they are car free, they may spend the money they save on overseas trips ... all of which makes inferences about density and emissions harder.

    report
    1. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Geoff Russell

      Geoff, the ACF atlas uses income as a major component variable linked to the Household Expenditure Survey. I agree - inferences about density are very difficult, indeed near impossible, which is why I argue it ought to receive less attention in planning debates.

      report
    2. Geoff Russell

      Computer Programmer, Author

      In reply to Jago Dodson

      Agreed. I was a little rushed in my first reading. I guess density gets seized on like
      all sorts of other things are seized on ... to divert attention from the big issues that
      nobody wants to deal with. We have to reforest 70 million hectares, totally restructure
      our energy system (nuclear being the only proven technology, with solar and wind failing dismally to be anything other than niche contributors), and totally change our
      diet. And that's just to get a fighting chance of getting CO2 back to 350 ppm by 2150.
      It's no wonder people clutch at "high density living", "reduced population" or other distractions.

      report
    3. Eclipse Now

      Manager of Graphic Design firm

      In reply to Geoff Russell

      Geoff,
      why is New Urbanism a distraction? My sister-in-law has a Phd in this stuff. I assume you're just mouthing off against New Urbanism because this isn't an article about the wonders of nuclear power. That's a rather ignorant false dichotomy, and it does nukies no favours to come across so superficially judgemental.

      If you want to go a LITTLE deeper, try this article by Alex Steffen at Worldchanging. "My other car is a bright green city".
      http://www.worldchanging.com/archives/007800.html

      report
    4. Geoff Russell

      Computer Programmer, Author

      In reply to Eclipse Now

      I apologise for the tone ... definitely too flippant. But not the content. People were spruiking what you are calling "the new urbanism" back in the 70s when oil prices spooked them and some people realised that cars were more trouble than they were worth. Lots of nice small projects proved how wonderful the concepts were, but emissions just kept rising. Sustainable long term co2eq emission level is about 1 tonne per person. Will new urbanism deliver that? Really, do the maths. Food alone on current diets will
      bust that limit even with NO cars, tiny houses, etc etc. CSIRO cited the current Australian diet at 6 tonnes per person per annum of co2eq... that's a bit high in my view, but you can see the problem. Rebuilding cities isn't CO2 neutral. We are stuck with a vast amount of infrastructure that won't go away.

      report
    5. Eclipse Now

      Manager of Graphic Design firm

      In reply to Geoff Russell

      Hi Geoff,
      Public opinion is already divided quite enough on climate change. If we can replace coal fired power with nukes, and start a New Urban and public transport revolution in the face of peak oil, I will be happy with that. Food is the *last* area I would nag everyone over, as people are already fed up with us 'greenies telling them how to live'. Give me an Australian public transport and energy system running on clean nuclear power, and I would be OVERJOYED!

      Agricultural practices will have…

      Read more
  5. Ross James

    Engineer

    Surely the starting point is to stop increasing our population. Most of our increase comes from immigration. As you population increases, we move more and more towards high density living. Is this really what we want?

    report
    1. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Ross James

      Population issues are far more complex than a simple 'more population = higher density cities' equation. An array of social, economic and institutional factors shape how population is accommodated across our national settlement pattern even before one considers consumption.

      report
  6. Jane O'Sullivan

    Agricultural Scientist at University of Queensland

    Excellent article, Jago.
    The main reason density has been promoted as the answer to everything is as an excuse for population growth and consequent windfall gains in real estate.
    Let's not forget that the planet is only interested in our total aggregate impact, not per capita. There are many ways to improve urban and building design, for environmental impact, liveability and productivity, but all are easier to achieve without population growth than with it.

    report
    1. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to Jane O'Sullivan

      Jane, I'm reluctant to embark on a debate about population as it wasn't my intent in the article. Perhaps another time. However per-capita consumption measures are helpful if comparing effects of different forms of urban development.

      report
  7. David Bennett

    Architect

    Thanks for the article.

    Its a little puzzling... why go to the trouble of looking at the data only to say "maybe we should abandon...density" as a metric. If there us a correlation between residential density and carbon emissions (there is), then that is very useful information. The 'middle-ground' density is the hint for a transformational way for that will enable our cities to evolve.

    But firsst one important thing to clear up: The ACF's Carbon consumption atlas is a great start, but until…

    Read more
    1. Jago Dodson

      Associate Professor and Director, Urban Research Program at Griffith University

      In reply to David Bennett

      David, I should acknowledge the ACF (rather Chris Dey et al's) data doesn't factor in density per se. But the discussion accompanying the atlas which is accessible via links and tabs on the website explains the relationship between higher density and higher consumption. Methodologically disentangling the effect of urban residential density from the consumption patterns of those who occupy the buildings is very difficult. As for Shaping Suburbia - an interesting resource but as with everything in this topic area care is needed to disentangle socio-economic effects from biophysical consumption.

      report
  8. David Arthur

    resistance gnome

    Thanks for a sensible discussion. Have you any comments on the notion of "food miles"?

    Detached single-family dwellings CAN be made to work, IF

    1) Food is grown in the back yard, (with composting).

    2) Relatively large amounts of roof space per person is given over to solar power and solar hot water.

    3) The house uses rainwater tanks for non-potable uses (provided low-energy methods are applied to delivering the water, eg solar/wind pumps, header tanks)

    4) The family members employ varied low-fuel use transport - walking, cycling, public transport where available.

    There is limited opportunity for denizens of medium and high density housing to invest the time and money for points 1, 2 and 3. Of course, denizens of low-density housing would have to behave as citizens, rather than simply as consumers, for any of these suggestions to be adopted. I, for example, haven't got around to adopting some of these suggestions; I'll get around to it.

    report
  9. Russell Hamilton

    Librarian

    David Arthur's comment describes what has happened in my suburb. Over the last 20 years the backyards have been sold off, so now there are two houses each on 500 sq metres. I see that many, many houses have 12 or 16 solar panels, and front yards have large, well built vegetable gardens. There are substantial water tanks slipped in between house and fence. On the other hand the nearest residential apartment blocks have not changed one bit in 20 years, so the 'sustainability' trend seems to be favouring the house on the block.

    I could walk to a major bus route in 5 minutes, ride my bike to a train line in 10. But as I can't stand public transport (or living without a garden and a private yard) I drive to work. If fuel becomes expensive I'll commute by motorbike. I think houses on small blocks can be at least as environmentally sustainable as grouped living - and without all the hassle I hear apartment dwellers have with their 'corporate body'.

    report
  10. Stable Population

    logged in via Facebook

    Dear Jago

    I won't try to turn this into a population issue either, but good contribution to the debate with solid evidence to support you. I guess I would support the point Jane O'Sullivan made, that regardless of the per capita impact of high/low density, population growth (all other things being equal) increases TOTAL impact.

    If we can also tackle 'all other things', like design, consumption, etc. We can really achieve something. We may even start living sustainably.

    William

    report
  11. Jim Wright

    Retired Civil/Structural Engineer, IT Consultant/Contractor

    There is room in our cities for all varieties of housing, whether low-, medium- or high-rise. It all comes down to their conformance to appropriate planning regulations. Thus, high-rise buildings are quite acceptable if they are distributed in such a way as to provide satisfactory living conditions, do not cause unpleasant experiences such as wind tunnelling and so forth. However, recent experience suggests that current regulations are either not well formulated or are easily circumvented. so if…

    Read more
  12. Patrick Sunter

    PhD Candidate at University of Melbourne

    Interesting article.

    I'll just make the comment that I was studying at UniSA at the time Perkins et al presented the referenced study on housing form and energy consumption: and was surprised that they didn't include Australia's primary (only?) inner-urban medium-density co-housing development in the study, Christie's Walk. Especially surprising since it's located only about 5km from the UniSA campus. Christie's Walk was explicitly designed with a goal of much better than average environmental performance, whilst still providing a high level of amenity to residents - see http://www.urbanecology.org.au/eco-cities/christie-walk .

    report