We can’t afford this house!
The construction bids are in and our project is slated to be 100 percent more expensive than what we had budgeted. How did this happen and more importantly, is there any remedy?
After making the final payment for our plot of land in January 2012, I knew I had to tackle a new type of challenge: to manage the design phase of a residential construction project. I was determined to do it as professionally as possible.
One of the first books I bought is the excellent What Your Contractor Can’t Tell You: The Essential Guide to Building and Renovating. We religiously followed the author’s guidelines to select our architect and general contractor.
After interviewing over a dozen architects in Medellin and Bogota, we came across the Frente a las Rocas house published in the Axxis magazine.
At that stage, I was leaning towards a modern architecture design based on the Case Study Houses in general and the Stahl House by Pierre Koenig in particular, with its elegant horizontal lines seemingly merging into the horizon via a daring cantilever.
We found an architect in Bogota, Ivan Baquerizo who was willing to listen to our ideas, with the sensibility and skills to turn them into a set of plans that were faithful to our wishes. We signed his contract in April 2012 and shared with him all the information we had gathered so far, ranging from specification text files to Pinterest boards and our budget, which we estimated to be around COP $1,300,000 per sq m of living space.
Most architect engagements in Colombia are governed by decree 2090 of September 13th 1989. It is worth noting that the agreement we signed had a clause pertaining to the budgeting exercise:
The preliminary budget is prepared based on the draft plans and developed and updated during the preparation of the final plans. It is delivered together with the architectural details.
Picking the contractors that would work on the soil survey, structural engineering, electrical engineering, plumbing and landscape design was relatively easy. After a few meetings we had selected several recommended firms with a good reputation in Medellin:
We did not hire a sustainability consultant, despite evaluating several options, since the Living Building Challenge was out of our reach due to the greenfield nature of our lot, while LEED for Homes certification wasn’t open for small residential projects when we contacted the USGBC in early 2012:
Unfortunately, LEED for Homes is unavailable for projects that are outside of the U.S. that do not have 50 or more units.
In the absence of any avenue for certifying our home, I felt I could just train myself in green building techniques, reading books like Green from the Ground Up, and working with our house design team to implement the improvements that made most sense.
Meanwhile, we collaborated closely with Ivan Baquerizo on the schematic design of our future home. Our main inspiration was Kengo Kuma’s Glass Wood House built in Connecticut. After several design iterations, we converged on a site plan and floor plan that we loved.
In June 2012, Consulcivil conducted a soil survey, taking samples up to 8.45 m deep in three different locations of the lot:
The report, delivered in July 2012, indicated that high compressibility silt soils had been discovered, with material resistance from 64 kPa (1336 psf, Pounds Per Square Foot) to 104 kPa (2172 psf). These values were in line with the content of Chapter 4, Foundations, of the International Residential Code for One- and Two-Family Dwellings:
As a consequence, Consulcivil recommended the following type of foundations:
Isolated footings interconnected with overhead beams (…). The base of each of the footing should rest upon three micropiles, manually excavated and staggered, with a length of 1.50 m and a diameter from 0.3 to 0.4 m.
Foundations would have to be excavated and concrete poured for all the built up areas designated in the plans and listed below with approximate area measurements:
- Public wing: 260 sq m
- Private wing: 150 sq m
- Reflecting pool: 105 sq m
- Swimming Pool: 75 sq m
- Garage: 60 sq m
- Total: 650 sq m
Compare this figure with 268 sq m of actual conditioned space planned: the foundations for volumes that aren’t habitable account for 59 percent of the total built area.
The plan below depicts on the left the location of each steel column for the public and private wings of the house, while on the right we can see the concrete footings and micro-piles for both wings, the reflecting pond and the swimming pool:
Notwithstanding this, following his visit to the lot our architect’s opinion was that building on a pier foundation would be too onerous and wasn’t a viable option given our budget. Because the soil survey report and the structural calculations were unintelligible to me I readily accepted his advice.
In August 2012 we held a day-long charrette with the whole design team as well as a general contractor and discussed multiple facets of the project, though no major emphasis was placed on budget compliance and value engineering.
Throughout the design phase I found it rather challenging to facilitate good communication between project participants. The team wasn’t cohesive or self-organizing. I ended up being the nexus of the project, driving the design process, chasing team members for deliverables and reviewing all drafts.
Meanwhile, our architect indicated that he wouldn’t be able to deliver a preliminary budget: instead we should leverage the general contractors for determining quantities of materials and costs. In a similar fashion the structural engineers assured us that any contractor is capable of accurately estimating the amount of construction materials required using the detailed plans provided.
We attempted to manage our budget risk by requesting lists of construction materials with unit pricing from the electrical engineers, plumbers, landscape architect, etc. We also received quotations for ancillary products, like the kitchen, the solar thermal system, the swimming pool, etc.
Yet one year into the design phase, we essentially had no idea what the overall cost of the house construction would be. We took a leap of faith and decided to nearly double our budget to COP $2,400,000 per sq m of conditioned space, committing absolutely all our savings to our dream.
In May 2013 our architect re-adjusted the cost of the contract, arguing that the initially agreed scope of 250 sq m had increased to 366 sq m, including parking space, technical room and workshop in the garage. Without a 47 percent increase to his fees the project would turn into a loss for him due to the sheer amount of details, changes and complexity of the project.
A second and final charrette was scheduled for August 2013 to review the final deliverables of all the parties involved in the project.
Two new general contractors were invited to share their views on the project and one of them suggested that we re-evaluate the opportunity of a LEED certification. We contacted the USGBC and learned that our project would be eligible for participation in the LEED for Homes international pilot program. We promptly applied through the USGBC Web page and paid the $225 registration fee.
Sustainability had taken a back-seat since the first design charrette and reviewing our designs we discovered multiple areas ripe for green building improvements. During Fall 2013 we hurriedly integrated as many sustainable materials and practices as possible into our project without impacting the architectural and structural plans.
Eventually, in December 2013 we were able to issue a formal request for proposal to the three general contractors that had accompanied us during the design process. They were given 3 weeks to put together their bids, with a detailed costs break-down and timeline, based on a Cost Plus Incentive Fee (CPIF) contract model.
All contractor questions were formally answered. They received weekly copies of the Q&A and we shared all quantities of materials supplied to us, hoping that the contractors would eventually converge to ideal estimates.
A common complaint during the bidding phase was that the architectural finishes were a moving target, which was true: we had received the specifications from the architect at the end of November, reviewed them and found many issues that were communicated back to Ivan Baquerizo. December is a slow month in Colombia, and we did not receive updates from our architect until January 2014.
Despite all our efforts, the bids received just before Christmas were incomplete and showed a wide range of unit price estimates for several key items in each of the 16 divisions. I painstakingly conducted a detailed bid review over the festive period and returned the unit price estimates to the contractors with dozens of comments, giving them another three weeks to fine-tune their bids.
At the end of January 2014 the final bids were in, and I knew I couldn’t afford the house we designed.
Averaging the unit price estimates of all three bids and grouping them into categories loosely based on the 16 divisions, we get the following break-down of costs:
As the load bearing structure for the green roof exclusively relies on steel columns and beams as well as glulam joists, the astonishing cost of the foundations becomes evident: 19.9% of the total construction cost is devoted to site and concrete works. To that we need to add a sizeable fraction of the costs for moisture protection of the concrete footings and slab.
Concrete works also exhibit the second highest standard deviation of all project divisions, reflecting the contractors’ uncertainty when it comes to estimating quantities of materials for the foundations:
The cost per square meter and square foot of the project is as follows:
- Per sq m of conditioned space: COP $4,894,401 ($2,447)
- Per sq ft of conditioned space: COP $454,705 ($227.35)
- Per sq m of space with foundations: $2,017,999 ($1,009.00)
- Per sq ft of space with foundations: $187,478 ($93.74)
The dollar cost estimates assume a currency exchange rate of USD $1 = COP $2000.
Moving on to soft costs, the General Requirements division covers the salaries of the permanent and part-time staff, including the resident engineer, assigned to manage the project for its total estimated duration of 5 months, which makes sense for a complex stick built home.
In addition to construction costs, we had to pay the following fees:
- Permit fees and taxes: COP $7,330,133 ($3,665)
- Legal fees: COP $3,885,000 ($1,943)
- Architecture and engineering fees: COP $46,827,054 ($23,414)
The last line item will suffer a significant write-off since the plans produced by our team cannot be built and will need to be redesigned.
According to the December 2012 figures published by Banco de la Republica, the average price per sq m for residential housing in Medellin was COP $2,300,000 ($107 per sq ft) or 53 percent cheaper than our project:
Surprisingly, real estate in Medellin is on a par with the price per sq ft of an average home sold in the US in 2013 according to the NAHB construction cost survey: $115 per sq ft, including construction cost, overhead and general expenses as well as profit. The breakdown is shown below:
Comparing with our own costs classification would be meaningless since our project adopted a different construction method, seemingly less cost-effective for an equivalent square footage than the wood framing commonly used in the US. Labor costs are also higher in the US.
Nonetheless, there’s a useful lesson to be learned here: assuming a 50% cost reduction goal, our sq ft cost would match the average prices in the US and Medellin. To build an exceptional house on an average budget we will need to accept trade-offs.
Is LEED certification worth it in these circumstances? I believe it is because the price of the green rating process will be between 1% and 2% of the total construction cost and it can be offset via partnership deals and intangibles like the goodwill such a flagship project may generate. More importantly, building green was a cornerstone of our project from its inception, so why not certify it anyway?
In the world of Agile software development, a retrospective is a continuous improvement exercise to identify at the end of a project phase (called a “sprint”) what the team should:
- Start doing
- Stop doing
- Continue doing
A retrospective implies that the team will stay together to work on upcoming sprints, thus applying the lessons learned. However, a single home project like ours is effectively a one-off prototype. This is why I believe a team-based retrospective isn’t feasible: we did not have a true team, just a loose group of professionals with no incentive to work on process or cost improvements. I’m the only stakeholder, besides the general contractor, with a long term interest to seek improvements that will yield a higher quality project.
My personal retrospective will address separately process oriented and cost related matters.
- Stop doing: halt this first design iteration and recognize it as a dead-end since it yielded a project blueprint that is astronomically expensive. We must recognize that we’re too late in the design phase to apply effective value engineering. We simply do not have the financial resources to fund such an expensive construction project, especially with no access to loans in Colombia. Furthermore, contemplating resale value, the current project cost is out of touch with the reality of the real estate market in Medellin. We must get back to the drawing board instead of constructing a castle above a financial abyss.
- Start doing: put budget at the center of the next design phase with an overarching objective to halve the cost of conditioned space, down to $1200 per sq m ($112 per sq ft). We worked hard to create a design we liked but neglected to consider its implementation cost. We need to work on value engineering for all aspects of the project to preserve its essence but slash costs.
- Start doing: apply Agile and Lean practices to maximize value creation and minimize waste. This is a tough challenge in Colombia since Agile and Lean run contrary to the entrenched habits of the construction industry. At the core, we need to select partners that are willing to give Agile values a chance and recognize the usefulness of individuals and interactions, customer collaboration and response to change, compared to the traditional waterfall design-bid-build model that produced magnificent detailed plans that we will have to trash, representing a terrible waste of time and money. Note that Lean very much applies to the construction phase as well.
- Start doing: work closely with companies that have a vested long term interest to see the project succeed. Going the extra mile only makes sense in that case. The general contractor is a prime example of a partner we need to treat differently. In retrospect, a competitive bidding process with set design guidelines wasn’t conducive to value engineering. We need to select a general contractor and work closely with him: the added cost of design interactions will be recouped during the construction phase, resulting in a win-win scenario.
- Continue doing: focus on design and planning. Several research studies confirm that improper planning at the pre-construction stage is the most significant factor on cost overruns in residential projects. To effectively manage budget and risk we must clearly define the scope of the project to minimize subsequent change requests.
- Start doing: split the project into two phases: 1) living quarters 2) pool, garage and landscaping. Tackling the whole project in one sweep would exceed our budget. Completing the living quarters first will allow us to move in faster, thus saving on rent and storage house expenses. We will need to carefully prepare for the transition from one phase to another, making sure that electricity and water meters are sited properly and a temporary solution is found for rainwater harvesting. All dependencies between phases will have to be evaluated.
- Stop doing: the cost plus incentive fee (CPIF) contract model will be simplified to a standard cost plus agreement that Colombian general contractors know and love. Besides creating some unease with the general contractors, I suspect the CPIF model encouraged them to use extreme safety margins while estimating costs in order to increase the likelihood of earning the promised cost reduction bonus.
- Start doing: invest wisely on LEED recommendations that are compatible with our LEED Silver certification goal, the project scope and our local conditions. For instance, we haven’t clustered together the bathrooms and the kitchen because we value privacy more than a totally efficient hot water distribution system. The loss of LEED points in that section can be compensated by installing PV panels that will lower our electricity bill. Likewise, installing an HRV or ERV is overkill in the mild climate of Medellin.
- Start doing: increase the percentage of foundation space devoted to conditioned areas of the house. There are many reasons why a single story home is less cost-effective than a two story home. Let’s not make matters worse by building expensive foundations for spaces that we don’t really need (e.g. eliminating the reflecting pool) or do not require a foundation (e.g. use a suspended wood deck, possibly covered with a thin layer of concrete, around the house walls instead of extending the concrete slab).
- Stop doing: as a corollary to the above, let’s eliminate the reflecting pool. It adds 105 sq m of expensive foundations while serving no functional purpose. We will have to re-visit the landscaping of the garden between the public and private wings as a result.
- Start doing: refactor the foundations from a structural engineering standpoint to save money while complying with the NSR10 norm. As it stands, the soil survey engineers prescribed the foundations for our house. Consulcivil is a fantastic firm but they aren’t structural engineers by trade: we need to perform a critical analysis of their recommendations to seek a more cost-effective type of foundations for our home. This may require contracting structural engineers with ample experience in seismically active regions like Japan.
- Start doing: consider building on stilts to avoid massive excavation work. Our existing design would force us to remove over 1300 m3 of soil to create a flat platform at 2149 m altitude. Besides the hefty expense, the environmental impact is not negligible. Building on stilts would cut on bulldozer work and help us achieve the ethereal beauty of the Glass Wood House: this would be a dream come true!
- Start doing: explore lightweight alternatives to the current house design. With a heavier house comes the need for stronger foundations and more extensive seismic bracing. The prime candidate for the chopping block is the green roof, which adds 120 kg of dead weight per sq m when saturated with rain water. The steel skeleton could also be optimized, preserving it but increasing the span of the steel beams between columns. However, from a green building perspective steel suffers from thermal bridging issues, while pre-fabricating to minimize on-site work may or may not be possible. A promising alternative is the glulam/plywood structure of the Aero House built in Japan.
- Stop doing: get rid of the green roof. Not only does this shave 120 kg per sq m, it also lowers by up to $70 per sq m the cost of the roof, which is a large expense in a one story home. Besides, we do not need to build the largest residential green roof in Colombia: we are already surrounded by 5083 sq m of Nature for our enjoyment. A single ply cool roof is almost equally earth-friendly and would allow us to catch 100% of the rainwater, directing it straight to our swimming pool, thus saving us the cost of an extra cistern. We just need to make sure that the roof is ready to receive our solar thermal domestic hot water system, as well as PV panels, with easy access to the equipment.
- Start doing: review the roof slopes — can we simplify them? Complexity costs money and the contradictory requirements to a) hide the roof b) try to keep a low slope (i.e. 2% or less) in order to avoid a “top heavy” facade c) drain the abundant rains that fall in that part of Colombia, pushed our architect to design a roof with multiple slopes that must be built with large dollops of mortar and plenty of manpower.
- Stop doing: lower the window to floor area ratio by eliminating windows that aren’t essential. LEED v4 for Homes, practically requires double glazed windows in order to earn a minimum of 1.5 points. The current house design includes 238 sq m of windows and needs to be reviewed to a) ensure the South-facing glazing area is at least 50% greater than the sum of the glazing area on the East- and West- facing walls b) substitute walls or other transparent or semi-transparent materials for double glazing in spaces where views are only occasionally enjoyed (e.g. the private wing corridor and the bridge in particular). Where windows are desirable, try to assemble fixed wood frames and double-glazing panels of standard sizes instead of buying ready-to insert, pre-framed, windows.
- Continue doing: pure structure-as-finish space. I’m quoting the fine engineers at Structured Environment who expressed our modern architecture inspired quest for simplicity and clarity of forms, visual expression of structure and truth to materials. Our goal is to promote building durability and beauty by channeling resources into better structural design and materials instead of expensive finishing touches.
- Start doing: design a material-efficient framing solution that can be largely prefabricated in order to reduce waste and shorten the duration and cost of work on site. In LEED v4 for Homes, the Materials and Resources section on material-efficient framing has changed, cutting down requirements and points from 5 to 2. Two points can be earned by combining advanced framing techniques and SIPs. Nevertheless, due to our difficult site access condition, it is worth evaluating pre-fabrication because of the resulting shorter build time. This needs to be balanced against the general contractor’s skills and capabilities so as to stay cost effective: familiarity with a building technique results in faster and cheaper work.
- Continue doing: use products that were extracted, processed, and manufactured locally or environmentally preferable products. This is another LEED v4 for Homes requirement for which we can earn up to 4 points. To reduce costs, we should strive to use local products whenever possible, except when there’s no local substitute to an item that needs to be imported (such as the EPA compliant wood burning stove).
Originally published at casaleed.org on February 20, 2014