By the Creating the Commons finance sub-committee · August 2026
A look at the financial implications for a library, a senior center, a city hall and a cultural center. Every number on the following pages comes from an invoice, an adopted budget line, or a model you can move yourself.
No site decision has been made. This is analysis for the Council and the public, not a recommendation of record.
Each of these is difficult to fund on its own. Together they can be financed.
10,000 sq ft today, and every study recommends roughly double that. On its own: a $24–34M rebuild serving one need.
Programs run across four addresses. On its own: a fundraising campaign and a relocation, likely outside the city.
Space is tight, and most staff already work out of Public Works. On its own: a similar capital cost, with no rent to help carry it.
In the flood zone, with $555,000 of flood and ADA work already adopted. Further flooding remains likely.
Sources: Sonoma County Library Facilities Master Plan (Nov 2023); City Ad Hoc report (May 2024); FY 2026-27 Adopted Budget, CIP 0714-26.03.
Five sites were studied before the shortlist. Here is what each offered, and why it did not move forward.
| Site | What it was | Why it did not move forward |
|---|---|---|
| Rite Aid 218 N Main | Downtown adaptive reuse, 15–20,000 sq ft | Either $10M to buy or an estimated $40,000 a month to rent — and a build-out of $5–10M on top of either route. Another buyer moved first. |
| Calder Creek or the High Street lot | City-owned property. Downtown civic commons with creek daylighting | Brownfield, with large CEQA and environmental concerns, Laguna flooding, and the Coastal Commission in the frame. It also carries the highest civic impact of any option studied, with creek corridors and a creek restoration already under way, which would be a real benefit to the city. Ranges from $40M to $60M, and the longest entitlement path in the study. |
| Railroad Forest | New construction, ~45,000 sq ft | ~$25M green-field build, with Laguna flooding, environmental impact and traffic flow all unresolved. A strong option, though slower and more expensive than buying an existing building. |
| Tow site 332 Petaluma Ave | 3,700 sq ft adaptive reuse | $1.5M to acquire, but it needs $3–4M minimum in improvements and possibly more — roughly $5M all in, for about a tenth of the space needed. It remains the lowest-cost option, and is best suited as an interim step. |
| Youth Annex | City-owned parcel | Kept for youth programmes. Not treated as saleable or rentable. |
Sources: Building the Commons site evaluations 2026; committee comparison matrix. Prices as documented at the time of evaluation and not re-verified since.
Option one
Waiting carries its own cost. The next page sets it out.
Every line is a paid invoice or an adopted budget item. Nothing here is an estimate.
| Sheet total, three city buildings | $279,660 |
| − Personnel (follows the people, not the buildings) | −$82,642 |
| = Facility cost, three buildings | $138,718 |
| + Library maintenance | +$11,000 |
| Every year, to change nothing | $149,718 |
Personnel is excluded on purpose. Staff costs follow the people, so a move would not save them. Leaving them out makes the do-nothing total lower, not higher.
The Cultural Center's $58,300 contribution to operations has been struck from this table. It has not been confirmed as facility-linked rather than programme funding, so it is left out until it can be sourced.
Ten years of it
On invoices and adopted CIP alone. $2.91M if the capital allowance is carried.
At the end of ten years, the city still owns the same four buildings.
Year two alone carries the Cultural Center's adopted $555,000 flood and ADA project. Operating costs escalate 4.5% a year.
Operating costs would not fall to zero at O'Reilly. One roof instead of four makes them smaller, not absent: utilities, insurance, cleaning and repair continue, and deferred maintenance and the adopted CIP follow the services wherever they sit. What this table measures is the difference between the two, and that difference is narrower than the $149,718 total suggests on its own. Timing matters as well. Construction costs rise each year, so the same project generally costs more built later than built sooner.
Sources: FY24-25 paid invoices, three city buildings; FY 2026-27 Adopted Budget facility detail; CIP 0714-26.03. Ten-year totals recomputed after the $58,300 line was removed, on the same escalation basis as the earlier draft.
Option two
Alameida Architecture, May 2019 — the only fully costed library scheme anywhere in the record.
We budget $24M. The range is $24M to $34M, and the spread reflects how much of the 2019 scheme gets built rather than uncertainty about prices. Carrying $24M splits the difference on completion and gives the City a figure it can plan against.
Why the top moved. Bay Area construction escalation ran 6–10% a year through 2021–23. Comparables: Half Moon Bay, 22,000 sq ft for $36M in 2018. Walnut Creek, 30,250 sq ft for $33M in 2012. Santa Cruz's downtown branch is the closest current regional comparable, and a sourced figure should be added here before it is quoted.
What $25.1M bought: 21,600 sq ft over two storeys — a net gain of 11,600. Expanded reading room, children's and teen zones, meeting rooms, staff areas, rooftop solar, new steel frame. $1,164 per square foot, all in.
No parking count and no parking cost appears in the 2019 study, or in any City document for this site. The committee's site review recorded parking on this block as a crisis.
Two questions for the committee: does the 2019 scheme still meet parking requirements, and how is that work funded?
Verify the LANTERN costs. This is an action item for the committee. The May 2019 Alameida estimate is the only fully costed library scheme in the record, and it has not been re-priced since — neither the scope, the unit rates, nor the escalation applied here. Before any figure on this page is carried into a budget, a cost estimator should re-base the estimate against current bid results.
Plan D
Refurbish and partially expand the existing library rather than replace it. It has not been costed anywhere in the record. It is also the option most likely to bring a rebuild within the $24M budget, so it deserves a place alongside the three above, and a real estimate.
Source: Alameida Architecture, Expansion Feasibility Study (May 2019) — $25,142,400 over 21,600 SF. Escalation band from Bay Area cost index history 2019–2026. The $24M–$34M range is the sub-committee's own working band, not a figure from the 2019 study.
Option three
All-in cost divided by the finished square footage each option delivers. Lower is better.
Buying a building that already exists is cheaper per square foot than building a new one, because the concrete, steel, roof, wiring and parking lot were paid for in 2001.
Sources: LoopNet listings for 1003, 1005 and 1007 Gravenstein Hwy N; Keegan & Coppin brochure; Alameida Architecture (2019). The downtown range is the committee's construction-cost basis; the LANTERN row is the study's own all-in basis — they are not the same measure and are shown separately for that reason.
From a Council decision to people walking in.
A rebuild finishes somewhere between 2031 and 2034. The O'Reilly option opens in 2028.
That difference of three to six years matters to the residents waiting to use these services.
Rent and increment set against what the loan costs each year. Every setting below is an assumption, not a commitment.
Loan: IBank ISRF 4.39%, 30 years, level payment with every surplus dollar applied to principal. Rent income is the leased area times the rate set above, less 22% for vacancy and operating cost, rising 2.5% a year. Both are sliders: the default is 40,000 sq ft of the 90,258 sq ft campus at $1.75, which is $840,000 gross and $655,200 net in year one. The rate is a placeholder pending a written broker opinion, and either slider can go to zero. The private-use line is the one to watch. Tax-exempt borrowing allows roughly 10% private business use — about 9,000 sq ft here — and the default lease is more than four times that. Space is a rule of thumb rather than the legal test, which turns on the financed portion and the revenue from it, and bond counsel has not ruled. If the borrowing has to be taxable the rate rises and every figure on this page moves with it. Year 1 = FY2028-29.
The same model at three settings. These follow the sliders on the previous page.
All three start from the same base: an all-in cost of $18.0M, $11.0M borrowed, and $666,594 a year in debt service at the current settings. Each lever moves from that same starting point, so the columns below are directly comparable.
The first two levers change what the city borrows. The third changes when the debt ends.
Rent covers the payments. The increment can retire the debt early, if it is directed here.
Rent alone, increment used elsewhere
Loan cleared in year 21. Interest paid: $6.7M.
Rent covers the scheduled payment and little more. The loan runs most of its full term.
Rent plus the increment, redeployed here
Loan cleared in year 15. Interest paid: $4.7M.
The increment saves $2.0M in interest, and continues producing revenue after the debt is retired.
What the district can and cannot do. It cannot buy the building, because it does not accumulate $18M quickly enough. What it does well is repayment: rent carries the early years while the increment is still small, and the increment then retires the debt early. Formation is already complete, so the question before the committee is where the increment is directed — to the Commons, or to other priorities. These two boxes follow the sliders on page 11.
EIFD increment reconstructed from the district model's published anchors. At the default settings this reproduces the analysis of record — year 15 with the increment redeployed here, year 21 without. An earlier draft quoted year 17 and year 27; those figures came from a 1%-a-year rent escalation rather than the 2.5% used here and throughout this page.
Ticking a box with no figure behind it does not change the total. The chart shows where those gaps are.
Moves it up
Moves it down
The budget basis. The sub-committee carries $24M as the budgeted number for a downtown library rebuild, against an honest range of $24M–$34M — see page 7. The model above opens at +25% escalation with the adopted Cultural Center line ticked, which lands at about $32M. That is the figure the side-by-side on the next page uses, so the comparison with an $18M O'Reilly refurbish is like for like.
One line that cannot be ticked. Rental income: $0 — a 21,600 sq ft library has no leasable space; the whole building is the library.
Base: Alameida Architecture (May 2019), $25,142,400. Cultural Center figure is adopted CIP 0714-26.03. Debt service at 4.5% over 30 years. Items shown unpriced have no figure in any City, County or consultant document reviewed — they are named rather than estimated, and shown in gold on the chart so that a ticked box with no figure stays visible.
Apples to apples: a downtown rebuild at about $32M against an O'Reilly refurbish at about $18M. Both models sit at whatever you have set them to — move the levers on page 11 or 14 and these move with them.
These are not the same measure and should not be quoted as if they were. The O'Reilly rate is the cost of buying a finished building constructed in 2001. The downtown rate is the cost of designing and building new space today. The difference between them is real and worth making, provided it is described that way.
O'Reilly: $666,594 a year, with rent carrying a large portion of it and the increment retiring the debt early, and all four needs addressed.
Downtown: $1,963,486 a year from the General Fund, with no rent line and no increment, and three of the four needs still to address.
O'Reilly square footage is the confirmed campus total of 90,258 sq ft. Downtown is the 21,600 sq ft of the Alameida scheme. Rent of $1.75/sq ft/month remains a placeholder pending a written broker opinion, and the private-use question on tax-exempt debt remains unresolved — both would change the O'Reilly column.
No site decision has been made. The committee recommendation is due September 15.
Creating the Commons finance sub-committee · August 2026. Every figure on these pages traces to a named source or is marked as an assumption you can move yourself.
A note on geography
Every argument on the pages above assumes a shape for this town. It is worth setting out what that shape is, because it affects what we mean by downtown.
Sebastopol is not built around a square. It is built along a line. Highway 116 comes in from the north as Gravenstein Highway, becomes Main Street through the middle of town, and leaves to the south as Petaluma Avenue. Bodega Avenue — Highway 12 — crosses it and runs west into the county and on to the coast. Almost everything this town does happens within a few hundred feet of one of those two roads.
What we call downtown is a node on that line: the busiest one, with the most frontage and the most history. It is not the geometric centre of the town, and the street layout does not radiate from it.
Every one of the nine sites on the map below sits on the spine, or within a block of it. No shortlist could have done otherwise, because the corridor is where the town is.
First, distance here is measured along the corridor, not as a radius. The choice is not between downtown and somewhere else. It is between two points on the same road, about 1.1 miles apart.
Second, the library's catchment is not the downtown blocks. It is west county, and west county arrives along Bodega Avenue. For most of the people who use it, the trip is already a drive down Highway 12 — either way.
Third, and in tension with the first two: linear does not mean walkable. A mile along the spine is a short drive but a long walk, and the pedestrian and cycling experience on that stretch is the weakest part of the corridor. Improving it is a capital cost the City carries whichever site it chooses, and it is listed in the weaknesses on page 8.
1 O'Reilly campus — the leading option 2 Library & City Hall 3 Senior Center, 167 N High
4 The Legacy craft store 5 Cultural Center, 390 Morris 6 Rite Aid — ruled out
7 Calder Creek 8 Railroad Forest 9 Tow site, 332 Petaluma
Distances computed from geocoded coordinates. The map is schematic and roughly to scale. It is not a survey, and the road geometry is simplified to show the corridor and where each option sits on it. Creating the Commons finance sub-committee · August 2026.