Tessellar Blog

Monday, October 31, 2016

A Big Site

Large sites on the other hand are much more difficult to handle. One criticism of Honeycomb housing that is often raised is the concern about ‘getting lost in a maze'. It's true that winding roads can be disorienting. In small scale development, maintaining a sense of location is easy. It’s the big sites that pose a special problem especially if the issues pertaining to scale and monotonous repetition are handled badly. The problem of wayfinding is closely linked to the scale of a project and the extent to which the housing products are repetitive.

Even large terrace housing layouts can be a problem to navigate if the houses all look the same and road junctions are indistinguishable from each other.


Figure 10 A 130 Honeycomb proposal in Tangakak, Johor


A proposal in Johor provided an opportunity to show how this issue can be handled. Our client was a government linked company charged with developing a new township near Tangkak which was to be the administrative centre for a new District carved out from the Muar District in the north of Johor.
There was already a layout plan for a new township of about 400 acres of which 130 acres were designated as residential. Arkitek M Ghazali was asked to redesign the conventional layout that had already been prepared which comprised mainly terrace houses. The client wanted to assess if a Honeycomb housing was a feasible option. It was a difficult project in that the numbers of houses that were proposed was overwhelmingly high compared to existing demand. In addition, affordability was also a problem.

However, for us this exercise could serve as a test for how the Honeycomb layout could respond to problems of wayfinding and scale.

Figure 11 A Honeycomb layout pattern with small and large courtyards

One strategy adopted to overcome the problem of scale is the introduction of variety in the layout. A variety of cul-de-sacs have been designed. 


Figure 12 A variety of courtyards

There are large courtyards with pocket parks of about a quarter of an acre which are usually surrounded by quadruplexes and townhouses. There are medium sized courtyards that are surrounded by a lower density combination of quadruplexes and sextuplexes. There are also small courtyards which are surrounded by triplex and duplex units.


Figure 13 A variety of courtyards symmetrically laid out along a spine road

On top of these there are courtyards that open up to the main distribution roads and they can vary in size to give each one a unique identity. Mainly, these half-courts contain a combination of high-end duplex, triplex and detached houses. Variety along the winding main roads makes each junction more recognizable from the next. On straight roads we see only the road disappearing into the horizon, but prominent buildings where the road bends can become landmarks that give each stretch of the road its own identity.

Solving the Problems of Scale and Wayfinding


Figure 14 Main roads in the Honeycomb layout

The Honeycomb Housing Layout for this proposal builds on the “organic” concept that has been adopted by the original Planner for the overall layout. Three housing precincts are proposed with names based on the traditional practice of chewing the betel nut: Kampung Sireh and Kampung Gambir on the northern section and Kampung Pinang to the south.

* "Lingkaran Kampung Sireh" (Betel Leaf Village Ring Road) which circles an existing hill.
* "Jalan Kampung Gambir" (Gambier Village Road) which connects Lingkaran Kampung Sireh to the main 100’ feet road.
* "Jalan Kampung Pinang" (Betel Nut Village Road) which curves through the middle of the southern portion of the land.

Branching off from these roads are residential areas which comprise houses that are clustered around courtyards and cul-de-sacs. The dendritic, tree-like, hierarchy of roads allows the numbering of the roads in an easy logical manner that helps people find their way around. 


Figure 15 An example of road map to a house

This 50 foot wide boulevard, Jalan Kampung Pinang, that gently curves through the site will have a procession of different types of houses and a variety of small parks and amenities along each side of its whole length, such that the features that appear on the left hand side will be reflected on the right side. 

Figure 16 View of a spine road

The effect in to create a complex symmetry that mimics the symmetry found in all forms of life.

Figure 17

Figure 18

Figure 19

The houses will all feature Malay architecture with details that will build on the features found on traditional kampong houses found in and around Tangkak.

The Problem of Affordability

The other, certainly much bigger issue, was affordability. In Tangkak, 43% of the working population earns between RM1,000 to RM 2,000, and 28% earn between RM2,000 and RM4,000. As such, it is advantageous to provide for some housing priced below RM150,000, in particular, to compete in the single storey house category.  For this, we have introduced the townhouse type of housing, priced at RM125,000 where the sextuplex unit is further divided into upstairs and downstairs units which belong to separate owners. 
Figure 20 The Sextuplex Townhouse
Figure 21 The Sextuplex Townhouse floor plans


The quadruplex was proposed as the most common type of house, and it was priced at RM166,000 for the typical unit. Taking into account edge units (the Honeycomb equivalent of corner terrace houses which have extra land), the average price is RM182,000. This was lower than the proposed pricing of the terrace houses at RM188,000 in the original proposal.

Figure 22 The Quadruplex House
More expensive sextuplex and semi-detached house types were also included in smaller numbers.
Figure 23 The Sextuplex House
Figure 24 Semi-Detached House
The proposed development was in fact huge compared to Tangkak’s relatively small economy. Too huge. Even with the projected increase I industrial activity, it would have taken time for demand for housing to grow. The developer will have to launch small numbers of units at a time, especially in the initial stages.

As expected, the developer decided not to proceed with any of the plans for this new town.

Back to Table of Contents


Sunday, October 30, 2016

6.1 A Small Site in Merlimau

I’m often asked: “what’s the smallest piece of land that you need to be able to put in a Honeycomb layout. The answer is very small indeed.

The site is located in Taman Debunga, just at the edge of Merlimau, a small town in Melaka (figure 1).


Figure 1 Location

Triangular in shape, just about 1.3 acres, it is bounded by the back of a row of terrace houses on one side and high-tension pylons on another side. One enters from the third side which faced shops and terrace houses.


Figure 2 Overall view

The client here was a local developer with more than 30 years of experience, but one which was faced with what they viewed as a stagnating market. Although their main products – terrace houses – were becoming more and more expensive, they were not confident that raising prices was an option. Buying power in the towns that they were building was weak and there were competing developers who appeared content with thin margins.


Figure 3 Site layout

The client sought us out to help them overcome this problem. They had this small left-over piece of triangular land where they could experiment with a Honeycomb layout. The triangular site proved ideal for courtyard housing. We placed the houses around a cul-de-sac road. Only a pair of houses was allowed direct access from the existing road fronting the shops.

Experimenting with Variety

We managed to squeeze 14 units into 1.31 acres of land: 84% of the development land is sellable; road takes up the rest making up only 14% of the land used (figures 2, 3)).

The density of the Semi Detached houses is almost 11 units per acre, the same as that usually achieved by terrace houses. Small in-fill projects like this can be very efficient if they are able to piggy back on existing infrastructure like electrical connection, water supply and sewer lines and treatment plants as well as amenities like open green areas.

There were two basic types of semi-detached houses – single and double storey (figures 4 - 7).


Figure 4 Double Storey Semi Detach House


Figure 5 Double storey semi-detach house plans

Figure 6 Single-storey semi-detach house

Figure 7 Single-storey semi-detach house plans



Some of the double storey units were attached to double storey ones, but a few were attached to single storey ones, creating an interesting up and down roof-line (figures 8, 9).


Figure 8 Single and double-storey semi-detach house 
Figure 9 Single and double-storey semi-detach house
There were also variations in the position of the car porches, adding another level of complexity. Given the irregularity of the site, we could justify introducing a large variety of house-types in order to maximize the number of units.

Back to Table of Contents


Friday, October 28, 2016

5 Reducing the Cost of Infrastructure

We would reasonably expect that the amount of roads would be a good proxy for the cost of infrastructure : drains, water mains, sewerage pipes, electrical cables and telephone lines all run along roads – the shorter the road, the lower should be the length of the other services. To test out this out, we undertook to start comparing in detail the cost of water reticulation, sewerage piping, roads and drains in a honeycomb layout against that of a conventional layout.

REAL SITE COMPARISON

We approached Dato Eddy Chen, the CEO of a respected developer (Metro Kajang Bhd) to help us undertake a cost comparison between a rectilinear and honeycomb layout. He was sufficiently intrigued by the ‘honeycomb’ concept to participate in a research project to study the comparative cost of infrastructure of a ‘honeycomb’ layout.  A control layout was adopted. This was the layout of a recently completed 60 acre housing scheme in Kajang, Bukit Mewah.





Figure 1 Conventional rectilinear layout vs. Honeycomb in Bukit Mewah, Kajang

Given that Metro Kajang is a very experienced developer, a lot of effort had gone into squeezing the maximum revenue out of every single square foot. The layout, as far as rectilinear developments go, is highly efficient. The developer made available to us the full set of layout plans, engineering drawings, Specifications and Bills of Quantities. A “Honeycomb”layout was prepared. We then had an Engineering Consultant, Perunding Metro (no connection with Metro Kajang Bhd.) under the direction of Ir. C.T. Sia, prepare the Sewerage, Water Reticulation and Roads and Drains drawings using the details and specifications adopted in the original design. In this way, a fair ‘apple to apple’ comparison could be undertaken. The quantities of the main items were then measured and the results tabulated for comparison. Edmund Foo, of Quantity Surveyors, JUB Segar then provided realistic rates to apply against the quantities.

RESULTS

After completing the alternative honeycomb layout, a detailed breakdown of the land use was produced.  Similarly, the drawing of the existing layout was subject to the same detailed breakdown. The drawings of the original and honeycomb layouts are given in figures 1 above. The detailed land-use breakdown of the two layout designs is shown in table 1.

Table 1 Comparison of land-use area breakdown










































The main results are shown below (table 2). The Saleable land in the honeycomb option is 23.92 acres compared to 23.21 acres in the original layout, an increase of 3.1%. 


Table 2 Summary of land-use area breakdown



Saleable Land (acres)
Road Reserve (acres)
Green (acres)
Amenities (acres)
Units
Original Built Option
23.21
14.42
7.46
8.01
304
Honeycomb
23.92
13.56
7.76
8.01
328
Increase/(decrease)
0.71
(0.86)
0.3
0
24
% Increase/(decrease)
3.1%
(6%)
4%
0%
7.9%

The land to be used as road reserve in the honeycomb option is 13.56 acres compared to 14.42 acres in the original layout, a reduction of 6%. The Green area is increased by 4%, whilst there is no change in the area of the Amenities. The number of units of landed property is increased from 304 to 328 units, up by 7.9%.

WATER RETICULATION SYSTEM
The main cost-centres picked up are 150mm and 200mm UPVC pipes and 150mm and 200mm mild steel (ms) pipes. The mild steel pipes are used where the pipes run below the premixed roads. The layouts of the built and honeycomb alternative options are given in figures 2  and the comparison is shown in table 3. There is an overall reduction in quantities for the honeycomb option; however there are more mild steel pipes. 

Figure 2 Water reticulation system layouts

Table 3 Comparison of water reticulation system quantities


SEWERAGE SYSTEM
The main cost-centers picked up are 225mm and 300mm diameter Vitrified Clay Pipes (VCP) pipes and manholes. The layouts of the built and honeycomb alternative options are given in figures 3 (next pages) and the comparison is shown in table 5. There is an overall reduction in quantities for the honeycomb option. There is a 15% saving in the length of 225mm VCP pipes, a 37% reduction in 300mm diameter pipes and a 23% reduction in the number of manholes.

Figure 3 Sewerage system layouts

Table 4 Comparison of sewerage system quantities

DRAINS
The main cost-centres picked up are drains and culverts. The drains are 0.6m wide, 0.9m wide and 1.2m wide. The culverts are 1.2mX 0.6m, 1.2X 0.9m and 1.8mX 1.2m. The layouts of the built and honeycomb alternative options are given in figures 4 (next few pages) and the comparison is shown in table 6. There is an overall reduction in quantities for the honeycomb option. There is a 10% saving in the length of 0.6m wide drains, a 57% increase in the 0.9m wide drains and a 58% reduction in the 1.2m wide drains.  In addition, the length of the culverts were reduced – for the 1.2mX 0.6m, 25%; 1.2X 0.9m ,2%; and 1.8mX 1.2m, 31%.

ROADS
The total area of the premixed road surface for both options were measured as follows – 28,799sm for the original option and 26,668sm for the honeycomb option. Refer to figures 5 and table 7 again. The honeycomb option saved 2,131sm of road surface. 


Figure 4 Roads and drains layout

Table 5 Comparison of roads and drains quantities



OVERALL COST COMPARISON

We found that the honeycomb layout produced lower cost figures for all the works compared with the conventional design (table 6 and 7).

For the Water Reticulation works, there is an overall reduction in quantities for the honeycomb option, however there are more mild steel pipes. Since the mild steel pipes are more expensive, overall, the cost of the honeycomb project is only cheaper by RM2508. I.e. only a 1% reduction in cost.

For the Sewerage works, there is an overall reduction in quantities for the honeycomb option. There is a 15% saving in the length of 225mm VCP pipes, a 37% reduction in 300mm diameter pipes and a 23% reduction in the number of manholes. Overall, the cost of the honeycomb proposal is cheaper by a substantial amount - RM126,941, i.e. a 19% reduction in cost .

For the Drainage works, there is a 10% saving in the length of 0.6m wide drains, a 57% increase in the 0.9m wide drains and a 58% reduction in the 1.2m wide drains.  In addition, the length of the culverts were reduced – for the  1.2mX 0.6m, by 25%; 1.2X 0.9m ,by 2%; and 1.8mX 1.2m, 31%. Overall, the cost of the honeycomb project is cheaper by a substantial amount – RM181,500, i.e. a 13% reduction in cost .

For the Road works, the honeycomb option saved 2,131sm of road surface.  The cost of the honeycomb proposal is cheaper by a substantial amount – RM102,301, i.e. a 7.4% reduction in cost.

The cost of the original design was RM3,688,618 compared to the honeycomb design, RM3,275,368. The honeycomb proposal turned out to be RM413,250 cheaper, a substantial 11% saving. However, there are more units in the honeycomb version – 328 instead of 304. This is an increase of 7.9%. Taken on per-unit basis, the savings in water reticulation, sewerage, roads and drains are as follows:

Table 6 Summary of cost comparison



Total Cost (RM)
No of Units
Cost per unit
(RM)
Original Built Option
3,688,618
304
12,133.12
Honeycomb
3,275,368
328
9,985.88

Table 7 Detailed cost comparison



DISCUSSION
It is interesting to note that the savings in the total area of road reserve is quite small. This amount is certainly less than that of the examples given earlier. I believe the reasons for this are as follows:

1     There are more units resulting in a higher density. These extra units bring down the area of road reserve per unit, and seen in this perspective, the reduction in roads becomes higher.
2    Only about a third of the layout is taken up by terrace houses. As shown earlier in the comparison between 3 and 8 units of detached units in a rectilinear layout versus the ‘honeycomb’ layout, there is a reduction in the amount of savings in roads. This is due to the absence of back-lanes in detached housing. In addition, the rectilinear layout of the detached lots employed a lot of cul-de-sacs which are land-use efficient.
3       The land designated for ‘future development’ was obviously shaped to take on continuing rows of terrace houses. A ‘honeycomb layout’ would have sliced out a different shape. As it is, the layout of ‘honeycomb’ houses at the periphery of the land for ‘future development’ became rather inefficient.


Given the small saving in road reserve area, the results become all the more remarkable. The roads pavement area itself saw a reduction of 7%.  Quite substantial reductions were found for the cost of drains and culverts, sewerage pipes and manholes, and water mains.
   
Applying realistic rates to the outline bill of quantities, we found that the overall savings in the cost of the infrastructural services to be approximately RM413,250 or 11%, and the per-unit savings to be 17%.

Another factor to consider is the extra saleable land made available in the honeycomb layout. The extra land that can be sold is 0.71 acres or 30928 sf. If we price this land at RM50 per sf, the extra sale value unlocked will be RM 1,546,380.


CONCLUSION

From the studies undertaken above, we have found that the adoption of the ‘honeycomb’ layout has brought about clear improvements in land-use efficiency through a reduction in the area of roads. In the example, the reduction in the amount of roads, bring with it clear savings in the cost of roads, drains, water reticulation and sewerage lines. This example is, of course, based on a specific built project, and we cannot conclude that the employment of the ‘honeycomb’ layout will always give better results compared to rectilinear layouts. However, it has been shown that even in a case where the savings in amount of land used for road reserve is quite small, there is still substantial savings to be made in the infrastructure.

The reduction in roads brought about an increase in the saleable area. In this case study, the increase in saleable area is marginal, i.e. 3%. However the estimated contribution to the bottom line is a hefty RM 1.5 million, almost four times the estimated savings from infrastructure.

Certainly, the honeycomb technique has been shown to have its inherent advantages, and for developers tired of looking at rectilinear proposals, an honeycomb concept now presents a cost saving alternative.  More than that, the benefits to be made from increasing sales revenue through having more land to sell, can be substantial. 

Back to Table of Contents




Thursday, October 27, 2016

4 Reducing the Cost of Land

A key reason why the the developer of Nong Chik Heights agreed to experiment with Honeycomb housing was that we were able to convince him that our proposal was not only better in terms of quality, but that it was also cost-effective. In this  section, we will address the issue of cost.

TERRACE HOUSING

Terrace housing has long been considered the densest form of landed property development possible. Indeed, of all the types of housing in Malaysia, it is the terrace house that predominates . The typical lot varies from 16’ x 50’ to    24’ x 100’, but the most common lots now are 20’ x 65’ and 22’ x 70’. The ubiquitous terrace house plan has been designed   and re-designed many times but always within the same restrictive framework without much scope for innovation.

The housing layout has also become stereotyped. In the typical estate, the terrace houses are lined up along grid-lines with 40’ service roads in front and much narrower back lanes and side lanes. Communal areas for schools, civic and religious buildings, as well as open areas for children’s playgrounds and parks, are also provided. Despite the infrastructure provided, it can be said that the design of many housing estates does not really meet the practical needs of the average resident. 

Apart from the aesthetic boredom of rows and rows of houses, among the drawbacks of the terrace house layout is the lack of public security and any genuine sense of community. With the rising price of land in urban areas, many people are resigned to apartments. The terrace house, for all its drawbacks, has been elevated to the status of a dream-home.

HONEYCOMB HOUSING

In “Honeycomb Housing”, instead of rows of terrace houses, we are proposing that every house is in a cul-de-sac with a garden in the middle where giant shady trees will be planted. The courtyard in the middle of the houses is not just a street for transit: it is a place safe enough from speeding cars and strangers, for even pre-schoolers to play on.

Our aim is to recreate the best elements of kampong and small-town life: where children can play outside our homes with friends without fear from crime and traffic, in a community where people know and talk to each other. We are trying to create a more suitable environment for the “kampong boy of the future” – something better than our existing terrace houses. And honeycomb housing can deliver all the benefits of the cul-de-sac housing environment.

For all the benefits that Honeycomb housing may have over conventional terrace houses, it would be less likely to be adopted if the new layout was more expensive and less affordable. 
We will first deal with the issue of land-use here. The following post will address the issue of infrastructural cost.

SMALL-SCALE COMPARISON WITH TERRACE HOUSES

A honeycomb neighbourhood comprising 5   units of quadriplexes and duplexes is compared with a terrace house arrangement of an equivalent 5   units. Although the land size of the houses is the same, when we analyze the breakdown of land-use, we find that the area used up for roads (yellow) in the honeycomb layout is much less than that in the terrace house layout. Because we have made the green area to be of the same size, therefore there is more saleable land.

Figure 1  Five units Honeycomb vs. five units terrace houses


Table 1 Five units Honeycomb vs. five units terrace houses


A COMPARISON OF 5 UNITS

HONEYCOMB HOUSE
TERRACE HOUSE

(SM)
(%)
(SM)
(%)
ROAD
334
26
611
41
GREEN
93
7
103
7
HOUSE
861
67
761
52
TOTAL
1288
100
1475
100

We then compared a honeycomb neighbourhood comprising 16   units of quadriplexes and duplexes against a terrace house arrangement of an equivalent 16 units. It is demonstrated in the table shown that the honeycomb layout is more land-use efficient.


Figure 2 Twelve units Honeycomb vs. twelve units terrace houses



Table 2 Five units Honeycomb vs.five units terrace houses












A COMPARISON OF 16 UNITS

HONEYCOMB HOUSE
TERRACE HOUSE

(SM)
(%)
(SM)
(%)
ROAD
879
23
1323
35
GREEN
264
7
269
7
HOUSE
2721
70
2190
58
TOTAL
3864
100
3782
100

A similar exercise comparing 2 and 8 detached houses laid out in rows an against the same numbers of equivalent honeycomb houses comes to the same conclusion.


Figure 3 Two units Honeycomb detached houses vs. two units linear detached houses


Table 3 Two units Honeycomb detached houses vs. two units linear detached houses



HONEYCOMB DETACHED HOUSES
LINEAR DETACHED HOUSES

(SM)
(%)
(SM)
(%)
ROAD
334
26
426
33
GREEN
93
7
90
7
HOUSE
861
67
761
60
TOTAL
1288
100
1275
100

Table 4 Eight units Honeycomb detached houses vs. eight units linear detached houses

HONEYCOMB DETACHED HOUSES
LINEAR DETACHED HOUSES

(SM)
(%)
(SM)
(%)
ROAD
879
23
818
25
GREEN
264
7
235
7
HOUSE
2721
70
2190
68
TOTAL
3864
100
3782
100



Figure 4 Eight units Honeycomb detached houses vs. eight units linear detached houses


















MEDIUM-SCALE COMPARISON WITH TERRACE HOUSES

We then made a comparison between a ‘honeycomb’ layout comprising 258 three-bedroom low-medium cost double-storey houses of 1200sf built-up area on 15.6 acres of land, and that a terrace-house layout consisting of 288 equivalent 3 bedroom low-medium cost double storey houses of also 1200sf with 18’ frontage on 21.74 acres. Both layouts are efficient are theoretically efficient with the land size and shape suited to the requirements of the retilinear and honeycomb geometry. The size of land and number of units are not exactly the same, but this is acceptable because we are interested in the ratios.


Figure 5 258 terrace houses on 15.6 acres
Table 5 Comparing the terrace houses above with Honeycomb houses below













HONEYCOMB HOUSE
TERRACE HOUSE

(%)
(%)
ROAD
33
47
GREEN
9
9
HOUSE (Sellable area)
58
44
DENSITY (No of Units per acre
15
15
AVERAGE LOT SIZE
1658
1261


Figure 6 288 Honeycomb houses on 21.74 acres




























We find that ‘honeycomb’ housing produces here greatly increased land use efficiencies. These advantages are summarized in the mathematical table comparing the terrace housing   against quadruplex/sextuplex  honeycomb housing. The density is the same but the amount of road for the ‘honeycomb’ is only 33% against 47% for the terrace. Consequently, the average size of each lot is 30% larger!

REAL SITE COMPARISONS

We have done several comparative studies, comparing actual terrace-house layouts to alternative ‘honeycomb’ layouts,  to illustrate how honeycomb layouts are more efficient than conventional rectilinear grid layouts. We have done several comparative studies to illustrate how honeycomb layouts are more efficient than conventional rectilinear grid layouts. The study of alternative layouts at Demak Laut, Kuching in Sarawak iis one example. 
Figure 7 Terrace house layout in Demak Laut site



Figure 8 Honeycomb house layout in Demak Laut site

Table 6 Comparing the terrace house with Honeycomb house layouts in Demak Laut


HONEYCOMB HOUSE
TERRAC HOUSE

(%)
(%)
ROAD
33
47
GREEN
9
9
HOUSE (Sellable area)
58
44
DENSITY (No of Units per acre
15
15
AVERAGE LOT SIZE
1658
1261

In this example, there is the same number of units. The green areas and provisions for amenities are about the same. The terrace alternative yields only about 40% sellable residential land. This yield is quite common for any landed property development. However, the honeycomb layout can yield about 56% saleable land. The reason for this can be seen in the reduction in road reserve – from 38% to 23%.

Another example, for a project in Sungai Lunchoo, Plentong in Johor Bahru, shows again how the ‘honeycomb’ layout reduces the amount of road and improves the ratio of saleable land.


Figure 9 Honeycomb house with terrace house layouts in Sungai Lunchoo
Table 7 Comparing the Figure 9 Honeycomb house with terrace house layouts in Sungai Lunchoo





HONEYCOMB HOUSE
TERRACE HOUSE

(%)
(%)
ROAD
35.2
41.2

Fewer Roads
GREEN
10.9
7.6

More Green Spaces
HOUSE (Sellable area)
43.3
40.7

Larger Sellable Land
NO OF UNITS
224
224

Same number of Units


WHY IS THE HONEYCOMB LAYOUT EFFICIENT?

First of all, the back-lane in the terrace house situation is wasteful – this feature is totally eliminated in honeycomb housing.

Secondly, we can reduce the amount of circulation space in a through road by cutting it off at the end. The final length is replaced by paved area designed for turning.


Figure 10 Cul-de-sac vs. through road


















Given a fixed area and number of houses to access, the shorter the cul-de-sac, the less the area taken up by the road. A square cul-de-sac neighbourhood has less road area than a long rectangular one. A circular one by itself would be the most efficient. However, as shown below the circle does not tessellate.


Figure 11 Cul-de-sac layout of a square piece of land vs. a circular one







Figure 12 Circles do not tessellate




















However, hexagonal neighbourhoods interlock without gap or overlap.

The third consideration is the length of the distribution roads that encircle a precinct. The perimeter of a hexagonal precinct is 7% shorter than the perimeter of a square one of the same area.




Figure 13 The perimeter of a hexagon is shorter than that of a square
All of the above factors combine in honeycomb housing to produce greatly increased efficiency of land use. A terrace can be seen as a row of houses surrounded by roads. In contrast, honeycomb houses surround the road. It is easy to understand intuitively that roads accessing internally are more efficient than roads accessing houses from the external boundary. This accounts for the efficiency of cul-de-sacs in general, and partly explains the efficiency of ‘honeycomb housing’.

Back to Table of Contents