Committee on Design

This message has been cross posted to the following Discussion Forums: Regional and Urban Design Committee and Committee on Design .
-------------------------------------------

Zoning regulations produce random results within a built environment that must provide shelter for growing populations within sustainable limits. The definitions of dwelling unit and acre illustrate the point; but the point is applicable to the entire concept of zoning, which is a legal nod to a sprawling issue.

 

At the present time, dwelling unit quantity is found by multiplying the acres involved by the number of dwelling units permitted per acre; but "dwelling unit" and "acre" often lack adequate definition. This means that a dwelling unit can be any size and the number of acres can include the total land area, even when some of it is unbuildable. This can leave too many dwelling units squeezed on a "buildable area" that is a fraction of the total land involved; or it can leave a permitted number of units occupying too much space with too little remaining for the parking and open space required. This defeats the concept of density, which is to limit building mass and pavement on the land we occupy for shelter. Parking is often considered open space in these circumstances. If too many dwelling units remain, requests to reduce parking requirements are submitted and intensity continues to increase. This is encouraged because the realistic development capacity of land is rarely equal to the density permitted. Density is a product of mathematical relationships, however, and capacity can be forecast based on the buildable land available, the average dwelling unit area involved, the parking and loading required, the miscellaneous pavement anticipated, the open space expected and the building height permitted. These factors influence achievable density, and optional values can be entered in the design specification templates of forecast models to predict development capacity and intensity alternatives.

 

All specification templates begin by calculating the "buildable area" and "core area" involved or required, since these areas must be established to accurately forecast development capacity. Buildable area in these forecast models is equal to the gross land area GLA minus percentages for present or planned public rights-of-way RDA and unbuildable areas UNB such as, but not limited to, ponds, wetlands, ravines and unstable soil. In other words, BLA = GLA - ((RDA% + UNB%) * GLA). Core area CORE is the buildable land area available for building footprint BCA and parking area after percentages of the BLA have been subtracted for project open space S, miscellaneous pavement MSP and loading areas LDA. In other words, CORE = BLA - ((S% + MSP% + LDA%) * BLA). This is the area that determines realistic shelter capacity when combined with building height options. It is found after a conscious provision for open space S has been subtracted. In the past, S has been a function of setback requirements, but too many zoning ordinances permit setback areas to be used for parking. This makes the ordinance a legal game rather than a tactical method of achieving a strategic plan.   

 

Core area sets the stage for development capacity and intensity. (Development capacity means gross building area GBA per buildable acre. Intensity is project open space per 1,000 sq. ft. of development capacity.) It's not hard to imagine that a given land area will accommodate more dwelling units if the open space and parking requirement is less. This makes arguments over parking, setbacks and permitted dwelling units inevitable - when capacity and intensity cannot be accurately predicted. In these cases, open space is often sacrificed to accommodate unrealistic density expectations encouraged by impractical zoning ordinance provisions.

 

Within the core, multi-family residential development capacity is inversely proportional to the aggregate average dwelling unit area AGG provided -- when all other design specifications are held constant. Table 1 has been removed from the design specification template of forecast model RG1L to explain the AGG value. It shows how the average dwelling unit area AGG of a project is a function of the dwelling unit types DU, dwelling unit areas GDA, dwelling unit mix MIX and building efficiency (e) values chosen. Building efficiency is a factor because AGG is equal to the gross building area divided by the number of dwelling units provided. This means that building support areas must be part of the calculation. These support areas include, but are not limited to, circulation areas, mechanical rooms, exterior wall thickness, etc. The ratio of all net dwelling unit areas to gross building area is referred to as building efficiency. The comprehensive dwelling unit area CDA calculated in Table 1 increases each GDA value to account for the support areas anticipated by the building efficiency percentage (e) entered. The result is a forecasted AGG value. The AGG area permitted per parking space (a) is also calculated based on the parking requirement (u) entered at the top of the table. This also equals the gross building area permitted per parking space because of the nature of the AGG value.

 

TABLE 1  DEFINING AN AVERAGE DWELLING UNIT (AGG)

 

Parking lot spaces planned or required per dwelling unit					u=	1.5
Garage parking spaces planned or required per dwelling unit					Gn=	0
Building efficiency as percentage of GBA
					e=	80%	must have a value >0 entered
Bldg. support as fraction of GBA					Bu=	20%	e + Bu must = 1
Dwelling Unit Mix Table:
DU		GDA		CDA=GDA/Be		MIX					PDA = (CDA)MIX
dwelling unit type		gross du area		comprehensive du  area -  including support		du mix					Pro-rated du area
EFF		350		438		10%					44
1 BR		500		625		30%					188
2 BR		1,000		1,250		60%					750
3 BR		1,200		1,500		0%					0
4 BR		1,400		1,750		0%					0
AGGREGATE AVG. DWELLING UNIT AREA								(AGG) =			981
				GBA sf per parking space				a=			654
Enter zero in the adjacent box unless you wish to override the AGG value calculated above											0

 

 

The data shown in bold within boxes in Table 1 are design specification values that can be modified to explore options. Table 2 illustrates another MIX option. It should be clear from these tables that the AGG values simply represent an average of the dwelling unit mix proposed and building efficiency estimated. The building area (a) permitted per parking space increases in direct proportion to an increasing AGG value since the parking spaces required per dwelling unit remain constant.

 

TABLE 2  DEFINING AN AVERAGE DWELLING UNIT (AGG)

 

Parking lot spaces planned or required per dwelling unit					u=	1.5
Garage parking spaces planned or required per dwelling unit					Gn=	0
Building efficiency as percentage of GBA					e=	80%	must have a value >0 entered
Bldg. support as fraction of GBA					Bu=	25%	e + Bu must = 1
Dwelling Unit Mix Table:
DU		GDA		CDA=GDA/Be		MIX					PDA = (CDA)MIX
dwelling unit type		gross du area		comprehensive du area - including support		du mix					Pro-rated du area
EFF		350		467		0%					0
1 BR		500		667		20%					133
2 BR		1,000		1,333		80%					1,067
3 BR		1,200		1,600		0%					0
4 BR		1,400		1,867		0%					0
AGGREGATE AVG. DWELLING UNIT AREA									(AGG) =		1,200
GBA sf per parking space									a=		800
Enter zero in the adjacent box unless you wish to override the AGG value calculated above											0

 

 

A dwelling unit mix table is attached to the design specification template of every multi-family residential forecast model. A typical design specification template for the parking lot design category is illustrated by Table 3.. Normally, a dwelling unit mix table is part of a design specification and a complete set of values is used by embedded equations to predict dwelling unit quantity NDU as the number of building floors increase or decrease. In this essay, the dwelling unit mix table has been separated to show the two options included as Tables 1 and 2 and all values in Table 3 have been held constant for all examples presented.

 

TABLE 3  THE DESIGN SPECIFICATION TEMPLATE of FORECAST  MODEL RG1L

 

Given:	Gross Land Area in AC	GLA=	10.000			435,600	SF
Public/ private right-of-way & paved easements		RDA=	0.100	W as fraction of GLA		43,560	SF
Net Land Area		NLA=	9.000	NLA in acres		392,040	SF
Facilities and features to remain as fraction of GLA		UNB=	0.200			87,120	SF
Gross Land Area Reduction		X=	0.300	fraction of GLA		130,680	SF
Buildable Land Area Remaining		BLA=	7.000	acres		304,920	SF
Est. gross pkg. lot area per pkg. space in SF		s =	350	enter zero if no parking required
Parking lot spaces planned or required per dwelling unit		u=	1.5	enter zero if no parking required
Garage parking spaces planned or required per dwelling unit		Gn=	0	enter zero if no parking required
Gross building area per garage space		Ga=	0	enter zero if no parking required
No. of loading spaces		l =	0
Gross area per loading space		b =	0	SF		0	SF
Project Open Space as fraction of BLA		S=	0.300	fraction of BLA		91,476	SF
Private Driveways as fraction of BLA		R=	0.030			9,148	SF
Misc. Pavement as fraction of BLA		M=	0.020			6,098	SF
Loading area as fraction of BLA		L=	0.000			0	SF
Total Site Support Areas as a fraction of BLA		Su=	0.350			106,722	SF
Core development area as fraction of BLA		C=	0.650	C=Su must = 1		198,198	SF
Building efficiency as percentage of GBA		Be=	0.800	must have a value >0 entered
Bldg. support as fraction of GBA		Bu=	0.200	Be + Bu must = 1

 

 

Table 4 compares the results produced by the Table 1 and 2 dwelling unit mix options when the design specification values in Table 3 are held constant in the forecast model RG1L. The results show that a constant density requirement will not produce consistent results when the dwelling unit MIX and AGG value is not constant. The results become even more inconsistent when other design specification values are modified. In other words, density limits in a zoning ordinance produce random results without a complete set of dwelling unit mix and design specification values.

 

We annex more land to compensate for random leadership because we do not forecast capacity to limit waste and control intensity. Development capacity has been forecast with rules of thumb and left to chance by ambiguous, conflicting requirements. This must change if we believe that plans are needed to provide shelter for growing populations within sustainable limits that protect their quality of life.

 

TABLE 4  COMPARISON of AVG DU AREA (AGG) to DU CAPACITY (NDU)

 

NOTE: NDU increase declines as floors (f) increase - when all design specifications are constant - because parking lots must expand to serve building area and dwelling unit increases. NDU increases are marginal above 5 stories because a building footprint finds it increasingly difficult to shrink in response to an increasing parking lot within core area limits.

 

PKG LOT SOLUTIONS ABOVE 5 STORIES ENCOURAGE OPEN SPACE SACRIFICE & PKG VARIANCE REQUESTS

 

 

 

Table 5 indicates that GBA efficiency (e) further destabilizes the results produced by an isolated density requirement. It shows that the same one story building could yield from 108 to 142 dwelling units depending on its efficiency when the design specifications of Table 3 are held constant and the dwelling unit mix of Table 1 is considered. Unfortunately, parking requirements increase as efficiency increases -- and open space is often sacrificed in response.

 

TABLE 5  COMPARISON OF BUILDING EFFICIENCY TO NDU CAPACITY

 

NOTE: NDU increase declines as floors (f) increase - when all design specifications are constant - because parking lots must expand to serve building area and dwelling unit increases. NDU increases are marginal above 5 stories because a building footprint finds it increasingly difficult to shrink in response to an increasing parking lot within core area limits.

 

PKG LOT SOLUTIONS ABOVE 5 STORIES ENCOURAGE OPEN SPACE SACRIFICE & PKG VARIANCE REQUESTS

 

a	b	c	d	f	g	h	j	k	l
	e	e		e			e
	60%	70%		80%			90%
	AGG	AGG		AGG			AGG
	1,308	1,121		981			872
	a	a		a			a
	872	748		654			581
floors (f)	NDU	NDU		NDU			NDU
			A=1-(b / c)		A=1-(c / f)	B=1-(f / b)		A=1-(f / j)	B=1-(b / j)
1.00	108	120	10.2%	132	8.5%	17.8%	142	7.2%	23.8%
2.00	168	183	7.9%	195	6.5%	13.9%	206	5.4%	18.5%
3.00	206	221	6.5%	233	5.2%	11.3%	243	4.3%	15.1%
4.00	233	246	5.5%	257	4.4%	9.6%	267	3.5%	12.8%
5.00	252	265	4.8%	275	3.7%	8.3%	283	3.0%	11.1%
6.00	267	278	4.2%	288	3.3%	7.3%	296	2.6%	9.8%
7.00	278	289	3.8%	298	2.9%	6.6%	305	2.3%	8.8%
8.00	288	298	3.4%	306	2.6%	5.9%	313	2.1%	7.9%
9.00	296	305	3.1%	313	2.4%	5.4%	319	1.9%	7.2%
10.00	302	311	2.8%	318	2.2%	5.0%	324	1.7%	6.6%
11.00	308	316	2.6%	323	2.0%	4.6%	328	1.6%	6.2%
12.00	313	320	2.5%	327	1.9%	4.3%	332	1.5%	5.7%
13.00	317	324	2.3%	330	1.8%	4.0%	335	1.4%	5.4%
14.00	320	328	2.2%	333	1.7%	3.8%	337	1.3%	5.0%
	T3m1e6	T3m1e7		T3m1e8			T3m1e9

 

 

Tables 4 and 5 should make it clear that independent zoning requirements produce unpredictable results. This essay shows that one component of a design specification can undercut the intent of another when uncoordinated, and more than one alteration can produce complete confusion when interrelationships cannot be predicted. We will not be able to avoid sprawl as long as we attempt to compensate for this arbitrary leadership with annexation.

 

If you believe that the Built Domain depends on the Natural Domain for survival, then the first step is to define the space required for these natural functions. The space that remains is the only space available for the artificial world of a Built Domain, and it must accommodate the expanding shelter, movement, open space and life support divisions of our built environment. Shelter protects us. Open Space feeds and calms us. Movement and Life Support serve us. Living within the Built Domain will require accurate forecasting of shelter capacity options, since we must avoid sprawl that will consume the face of the planet if unrestrained. Accurate prediction will permit lifestyle options and variety to be evaluated within this domain. Decisions can then be expressed in terms that will lead many toward one goal, since we must protect the survival and quality of life in two separate, unequal worlds.

 

At the present time the public speaks with emotion in reaction to city planning uncertainty. If I substituted the word "medical" for "city planning" in the preceding sentence, I could be speaking of plague in the Middle Ages. We are now faced with a similar challenge in a different time with the same potential consequences. We call the challenge "sprawl" and stare at symptoms, but are not convinced it's a disease. Some can imagine the problem, but their language and tools cannot reach to the heart of the issue. Emotion becomes a substitute for knowledge and visions provide leadership. It should sound familiar. It's a path we've followed throughout the history we remember.

 

City planning has reached the "vision" milestone. Within the Built Domain, the next steps require an improved language, a precise measurement system, a standard evaluation format and the ability to duplicate success while encouraging individual personality. This should also sound familiar, since we've called these steps the scientific method. In medicine, the goal has been individual survival and quality of life. In city planning, the goal is mutual survival and quality of life based on a sustainable relationship between the Built and Natural Domains. If the Built Domain has limits however, we must learn how to shelter growing populations without excessive intensity that threatens quality and survival.

 

I have no answers, but I have attempted to contribute a language, a classification concept, a specification system, a context evaluation format, measurement standards and forecasting software that can be used to predict options. These are strategic tools that can lead to tactical decisions with the power to reconcile random results.

 

There is no alternative to a policy of life within sustainable limits -- if survival is the goal. It is just a matter of how the limits will be drawn. Random acts of zoning are not equal to the challenge; but they have established an important precedent that represents command potential.

Zoning has not changed since the early 20^th century, but the threat has changed and technology must adjust to the intelligence required. This belief has led to the book and software noted in my biography, which have been the resource for this essay and my contribution to an effort that hopes to produce sustainable cities and design.
 -------------------------------------------
Walter Hosack
Author
Walter M. Hosack
Dublin OH
-------------------------------------------