Methodology and sources
The data, formulas, defaults and tests behind the PVMalta.com calculators, and what we could not source.
Last updated: 22 November 2026. All sources read on that date.
This page sets out how the PVMalta.com calculators work: the data, every formula, the defaults and where each one comes from, how we tested the model, and what it cannot do. Where we found no public source for a number we say so and ask you for it instead of guessing.
1. Solar yield
Yield comes from PVGIS 5.3, the European Commission Joint Research Centre's free solar model (radiation database PVGIS-SARAH3, years 2005-2023). We downloaded results for one point on Malta (35.90 N, 14.51 E) and one on Gozo (36.04 N, 14.24 E), for a 1 kWp crystalline-silicon system, free-standing, at tilts 0 to 50 degrees in 5 degree steps plus 60, 70, 80 and 90, and at 24 compass directions 15 degrees apart, with PVGIS's default 14% system loss.
For a roof between grid points we interpolate bilinearly, for the year and for each month:
The calculators take a compass bearing (0 = north, 180 = south); PVGIS uses 0 = south, -90 = east, 90 = west, so az = bearing - 180, wrapped to -180..180.
P is the system size in kWp. Y already includes PVGIS's 14% system loss (cabling, inverter, dust, ageing) and the temperature, angle and spectral effects PVGIS models itself. If you set another system loss L, the yield is scaled by (1 - L/100) / 0.86; PVGIS applies its system loss as a flat factor, which we confirmed against PVGIS runs at 10% (worst difference under 0.3%). S is any extra loss you add for nearby shading, which PVGIS does not see (it only uses the distant horizon).
2. Hour by hour, and self-consumption
PVGIS hourly output for 2023 (tilt 30, facing south, east and west) gives the share s[m][h] of each month's daily energy in each clock hour (Malta time, summer time included). Other directions are interpolated by azimuth between south and east or west; beyond east or west the east or west shares are used, and other tilts use the tilt-30 shares (the shape changes little; the monthly totals still come from the exact tilt).
C is your yearly use. We found no published hourly or monthly load profile for Maltese homes, so the month weights w_m are the days in each month over 365 (the production chart lets you enter twelve readings instead) and the hourly shape l[h] is one of three assumed shapes, in percent of a day's use per clock hour:
| Shape | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| flat | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 | 4.2 |
| daytime | 2.3 | 2.3 | 2.3 | 2.3 | 2.3 | 2.3 | 2.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 5.3 | 4.5 | 4.5 | 4.5 | 4.5 | 4.5 | 3.0 |
| evening | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 4.9 | 4.9 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 8.2 | 8.2 | 8.2 | 8.2 | 8.2 | 8.2 | 4.1 |
These shapes are illustrations, not data. Average-day modelling also smooths out cloudy days and day-to-day swings in use, so it tends to overstate what a battery can shift.
3. Battery
We step through the average day of each month, hour by hour, and repeat the day until the battery's charge at midnight settles:
eta is the round-trip efficiency relative to the AC surplus; the default 0.89 / 0.97 = 0.918 comes from the Tesla Powerwall 3 datasheet (89% solar-to-battery-to-grid, 97% solar-to-grid). Independent tests (HTW Berlin 2025) put battery-only efficiency at 88-98%. Battery output adds to SC and the charged energy comes out of export. The battery's benefit is the difference between the cash flow with and without it; break-even is the first year where the summed difference reaches the battery's net cost.
4. Electricity bill and savings
Enemalta bills by cumulative yearly bands (pro rata over each billing period). With band limits b_k and rates r_k (residential or domestic, 5% VAT included):
The eco-reduction for a primary residence is 25% for a single person under 2,000 kWh a year; for two or more people, 25% on the first 1,000 kWh and 15% on the next 750 kWh per person, lost if use exceeds 1,750 kWh per person. Enemalta's own wording does not say "per person"; we follow the REWS report's reading, so confirm yours with ARMS. Each self-consumed kWh is worth the rate of the band your use falls out of, not the average price. When you enter a bill instead of kWh, the calculator finds the yearly use whose bill matches it.
5. Export income
f = €0.15/kWh for 20 years (PV under 40 kWp, no grant, REWS FIT table). Exports above 1,600 kWh per kWp a year, after year 20, or with no tariff earn Enemalta's proxy market price p_proxy, which we could not find published: enter it if you know it, otherwise it counts as 0. Feed-in income is exempt from income tax and VAT for households.
6. Ageing, payback and lifetime
d = 0.5% a year by default (median of measured field rates, Jordan and Kurtz 2013); module warranties assume about 0.4% a year after a first-year drop of 1-1.5%. SC, X and A are recomputed every year. Tariffs stay at today's rates (e = 0) unless you set a yearly change; r is an optional discount rate. Inverter replacement is not included.
7. Grant or feed-in tariff
In 2026 the REWS grant and the feed-in tariff exclude each other: the grant requires that the PV has no active feed-in tariff allocation. The comparison is:
with the 2026 Option A terms g = 65%, G_max = €3,000, G_kWp = €645 per kWp for PV with a hybrid inverter, and 75%, €6,000, €600 per kWh for a battery. The call (Call 2026/RES) closed on 12 September 2026, and the 2026 feed-in approval period runs to 30 November 2026. No 2027 call or scheme was published on 22 November 2026; both sets of terms can be edited for a what-if. Sources: REWS 2026 scheme page, Guidelines for Applicants, REWS leaflet.
8. CO₂
F = 364 g CO₂/kWh (Enemalta fuel-mix disclosure for 2025, provisional; 380 for 2024): the direct emissions of the average electricity Enemalta supplied, interconnector imports included. It is not a marginal or life-cycle factor, so treat the result as an estimate.
9. Roof area and suggested size
Current 425-475 W modules (LONGi, Jinko and Trina datasheets) need 4.2-4.7 m² of module per kWp; the default is 4.44. Rows spaced for no shading at noon on 21 December need about 5.7, 6.7 and 7.6 m² of roof per kWp at 10, 20 and 30 degrees. That spacing rule is a design choice, not a regulation; parapet setbacks, access and water tanks come on top.
10. How we tested it
The calculators share one engine (engine.js). A test script runs it under Node.js against PVGIS results saved on 22 November 2026 for nine orientations that sit between grid points. We accept 2% on the year and 4% on any month. All nine pass:
| Island | Tilt | Facing | PVGIS kWh/kWp | Engine | Year error | Worst month |
|---|---|---|---|---|---|---|
| Gozo | 28° | 202° (PVGIS 22) | 1660.2 | 1656.6 | -0.22% | 0.44% |
| Gozo | 73° | 80° (PVGIS -100) | 907.4 | 905.9 | -0.17% | 1.05% |
| Malta | 12° | 290° (PVGIS 110) | 1385.7 | 1385.5 | -0.02% | 0.19% |
| Malta | 13° | 128° (PVGIS -52) | 1515.9 | 1514.2 | -0.12% | 0.25% |
| Malta | 27° | 200° (PVGIS 20) | 1617.6 | 1614.2 | -0.21% | 0.38% |
| Malta | 33° | 172° (PVGIS -8) | 1630.8 | 1626.8 | -0.24% | 0.48% |
| Malta | 47° | 147° (PVGIS -33) | 1531.3 | 1529.7 | -0.10% | 0.38% |
| Malta | 55° | 340° (PVGIS 160) | 635.5 | 642.6 | 1.11% | 3.11% |
| Malta | 7° | 250° (PVGIS 70) | 1473.1 | 1472.0 | -0.08% | 0.12% |
The largest error is a steep roof facing north-north-west, where little sun falls in winter. For roofs facing between east and west the error is under 0.3% a year. The same script checks grid points against the raw files, the loss scaling, the tariff bands and eco-reduction, the grant caps, the battery's energy balance and the payback arithmetic.
11. Limitations
- Two PVGIS points stand for each island. PVGIS cells near the coast differ by a few percent.
- PVGIS models free-standing panels; panels flat on a roof run hotter and make a little less.
- Average days per month hide cloudy spells and daily swings in use, which flatters batteries.
- Load shapes, the proxy export price, installed cost, battery price and battery ageing are not sourced and are left to you.
- Bands are applied to yearly totals; Enemalta applies them pro rata to each billing period, which differs slightly when use is very seasonal.
- Grants and tariffs change. The figures are as published on 22 November 2026.
12. Not sourced: your input or a labelled range
We could not find a primary source for these. The calculators ask you for them, or show the range and say where it comes from:
- Installed cost per kWp, battery price per kWh: no primary Maltese price source. User input, no default.
- Enemalta proxy market price for exports above the 1,600 kWh/kWp cap, after year 20, or with a 2026 grant: REWS says it is announced annually; no current value found.
- Hourly and monthly household load profile for Malta: none found (Enemalta/NSO). Tool defaults (flat by hour, by days per month) are assumptions.
- Average consumption: only derived (Eurostat 4,617 kWh per household; 3,273 per household meter) or the regulator's 3,500 kWh convention; no NSO survey average.
- Lower-loss variant 10%: project assumption, not a published figure.
- NZIA +5% (grant leaflet): basis (percentage points vs share of grant) not stated in REWS documents.
- Soiling in Malta: no Maltese measurement; range from Mediterranean studies (2-6%).
- Battery capacity fade over life: no independent field rate sourced.
- Row spacing criterion for flat roofs (no shading at noon on 21 Dec) and roof setbacks: design choice, not a rule.
- 2027 grant call and 2027 FiT scheme: not published on the retrieval date (2026 FiT approval window ends 30 Nov 2026).
- Eco-reduction: per-person reading of the 1,000/750 kWh tiers taken from REWS; Enemalta's own wording is ambiguous.
13. Every constant
The full list the calculators read, with status, source and notes. The same data is in constants.json.
| Item and notes | Value | Status | Source |
|---|---|---|---|
| Electricity tariff | |||
| residential bands eur per kwhApplies to a Residential Premises Service (someone registered at the address). Bands are applied pro rata to the days covered by each bill (REWS 2026 Monitoring Report, s.3.2.2(a)ii). | 0–2,000: 0.1047 2,001–6,000: 0.1298 6,001–10,000: 0.1607 10,001–20,000: 0.342 20,001+: 0.6076 EUR/kWh, incl. 5% VAT, cumulative annual kWh | verified | enemalta.com.mt |
| domestic bands eur per kwhDomestic Premises Service (dwelling with nobody registered, e.g. second home). Same band limits. | 0–2,000: 0.1365 2,001–6,000: 0.1673 6,001–10,000: 0.2023 10,001–20,000: 0.418 20,001+: 0.686 EUR/kWh, incl. 5% VAT | verified | enemalta.com.mt |
| service charge single phaseSame for domestic (Enemalta domestic page). | 65 EUR/year incl. VAT | verified | enemalta.com.mt |
| service charge three phase | 195 EUR/year incl. VAT | verified | enemalta.com.mt |
| max demand tariffOnly three-phase services above 60 A per phase. | 21.05 EUR per kW of maximum demand per year | verified | enemalta.com.mt |
| vat rateAlready included in all published rates; do not add again. | 0.05 | verified | enemalta.com.mt |
| excise dutyIncluded in the household tariffs (REWS 2026 Monitoring Report). | 0.0015 EUR/kWh | verified | rews.org.mt |
| eco reductionPrimary residence, Residential Premises Service only. Applied to the consumption charge, pro rata over the billing period. Enemalta text reads '25% on the first 1,000 units, 15% on the remaining 750 units'; REWS report reads these as per person. Confirm the per-person reading with ARMS. | single person: discount: 0.25 condition: annual consumption < 2000 kWh, on all consumption two or more: tier1: discount: 0.25 kwh per person: 1000 tier2: discount: 0.15 kwh per person: 750 condition: lost if consumption exceeds 1750 kWh per person per year | verified | enemalta.com.mt |
| tariff unchanged 2025'During the year under review there were no changes in the retail tariffs.' Enemalta pages last modified 6 Feb 2026. | true | verified | rews.org.mt |
| eurostat avg price band DC2026-S1 provisional. Cross-check only; the tools use the band table. | 2024-S1: 0.1272 2024-S2: 0.1303 2025-S1: 0.1261 2025-S2: 0.1296 2026-S1: 0.1268 EUR/kWh all taxes, households 2500-4999 kWh/yr | verified | ec.europa.eu |
| Feed-in tariff | |||
| fit no grantREWS FIT Scheme Table (updated 29 Apr 2026): approval period 1 Jan 2026 to 30 Nov 2026, PV >=1 kWp and <40 kWp, any location, no grant or other De Minimis support on the capital investment, S.L.545.27 as amended by LN 101 of 2026. The 2026 approval window ends 30 Nov 2026; a 2027 scheme was not published on rews.org.mt on the retrieval date. | 0.15 EUR/kWh exported | verified | rews.org.mt |
| fit scheme capacity 2026 | 8 MWp (12.8 GWh/yr) | verified | rews.org.mt |
| fit with grantOnly for new PV approved with a grant under the 2025 (and earlier) RES rounds. Not available with a 2026 grant: the 2026 grant requires no active FiT allocation (REWS 2026 scheme page). | 0.105 EUR/kWh | verified | rews.org.mt |
| export cap kwh per kwpApprovals from 1 Jan 2013: kWp x 1600, no other cap (FIT table and REWS FAQ 6). Excess paid at Enemalta's proxy market price. | 1600 kWh exported per kWp per year paid at the FiT | verified | rews.org.mt |
| proxy market pricePaid for exports above the cap or with no FiT. REWS says it is announced annually; no current figure found. Leave as user input. | none | not sourced | rews.org.mt |
| allocation validity monthsOne-time extension possible at the FiT available on the extension date; FIT table note 3: extensions to capacity not covered by a scheme at original FiT minus 0.5c. | 12 | verified | rews.org.mt |
| fit taxREWS FAQ 15. | income-tax and VAT exempt for households | verified | rews.org.mt |
| export optionsREWS FAQ 7-8, 16; switch via ARMS Form R. | sell all generation (pay all consumption) self-consume, export surplus | verified | rews.org.mt |
| net meteringNot available to PV approved after 10 Sep 2010; no switch from net metering to FiT since Aug 2015 (FAQ 18-19). | closed | verified | rews.org.mt |
| 2026 grant (Call 2026/RES) | |||
| statusCall 2026/RES (GN 597 of 2026). Options B and C closed to new Part A from 13 Jul 2026 (GN 1060 of 2026); whole call closed from 12 Sep 2026 (Government Notice of 4 Sep 2026). Leaflet date 'end of November 2026' superseded. No 2027 call on rews.org.mt on retrieval date. | closed | verified | rews.org.mt |
| pv hybrid inverterOption A, PV system with hybrid inverter. Leaflet misprints the per-kWp cap as '€645/Wp'. | share: 0.65 cap per system: 3000 cap per kwp: 645 EUR | verified | rews.org.mt |
| battery | share: 0.75 cap per system: 6000 cap per kwh: 600 EUR | verified | rews.org.mt |
| hybrid inverter option BLeaflet only. Option B open to existing PV connected for at least 6 years. | cap per system: 1800 cap per kwp: 350 EUR | verified | rews.org.mt |
| nzia bonusLeaflet: '+5% NZIA financial compensation'. Whether +5 percentage points or 5% of the grant is not stated. | 0.05 | not sourced | rews.org.mt |
| grant excludes fit'provided that the photovoltaic installation has no active feed-in tariff allocation'. | true | verified | rews.org.mt |
| eligible cost incl vatGuidelines: total expenditure is the sum of costs including VAT, less discounts, extended warranty and import duty. | true | verified | rews.org.mt |
| min size | pv kw: >1 battery kwh: >2 | verified | rews.org.mt |
| durability yearsSystem must stay installed, grid-connected and operating for 5 years after payment; 10-year system warranty required. | 5 | verified | rews.org.mt |
| install deadline months | 8 | verified | rews.org.mt |
| previous roundHistoric. REWS page is dated 13 Mar 2025 but its text describes the round opening 23 Feb 2024 (EUR 4.8m, ITB calls in 2024). Do not label these as 2025 terms without a 2025 primary source. | pv standard inverter: share: 0.5 cap: 2500 cap per kwp: 625 pv hybrid: share: 0.5 cap: 3000 cap per kwp: 750 battery: share: 0.8 cap: 7200 cap per kwh: 720 | verified | rews.org.mt |
| Grid CO₂ | |||
| enemalta 2025Provisional. Supplier disclosure under S.L.545.34: electricity supplied by Enemalta incl. interconnector imports (33.45%); on-site self-consumed PV not included. Same table in REWS 2026 Monitoring Report. | 364 gCO2/kWh | verified | enemalta.com.mt |
| enemalta 2024 | 380 gCO2/kWh | verified | enemalta.com.mt |
| Household consumption | |||
| households electricity 2024Eurostat energy balance, households final energy use. Whether rooftop PV self-consumption is included was not checked. | 1086.336 GWh | verified | ec.europa.eu |
| households count 2024 | 235.3 thousand | verified | ec.europa.eu |
| avg per household 2024 | 4617 kWh/yr | derived | ec.europa.eu |
| avg per household meter 2024Meters include second homes and vacant dwellings, so this is lower than per occupied household. | 3273 kWh/yr | derived | rews.org.mt |
| regulator referenceReference bill used by REWS (Eurostat band DC, 2500-4999 kWh). A convention, not a measured average. | 3500 kWh/yr | verified | rews.org.mt |
| PV losses | |||
| pvgis system loss defaultPVGIS system loss covers cabling, inverter (DC to AC), dust and long-term degradation. Temperature, irradiance, angle-of-incidence and spectral effects are modelled separately by PVGIS. | 14 % | verified | joint-research-centre.ec.europa.eu |
| lower loss variantAssumption chosen for this project: a new, clean, unshaded system with a modern inverter. Not a published figure. PVGIS applies system loss as a flat factor, so yield scales by (1-L)/(1-0.14). | 10 % | not sourced | — |
| pvwatts breakdownNREL PVWatts v5 manual (Dobos 2014, NREL/TP-6A20-62641) Table 6. PVWatts models the inverter separately (nominal 96%), so its 14% is not the same bucket as PVGIS's 14%. | soiling: 2 shading: 3 mismatch: 2 wiring: 2 connections: 0.5 light induced degradation: 1.5 nameplate: 1 availability: 3 total: 14.08 % | verified | osti.gov |
| soilingPVWatts default 2%; IEA-PVPS T13-21:2022: annual soiling loss 5.86% in Crete, mean daily 4.4% in Malaga, 'at least 3-4%' globally. No Malta measurement found. | low: 2 high: 6 % per year | range | osti.gov, iea-pvps.org |
| temperature maltaTemperature and low-irradiance loss modelled by PVGIS. Same files: angle-of-incidence -2.59/-2.62%, spectral +0.46/+0.50%, total incl. 14% system loss -21.03/-20.90%. Roof-mounted modules with little ventilation run hotter than PVGIS's free-standing case. | malta: -6.17 gozo: -6.02 % of annual energy, S 30 deg, free-standing | verified | PVGIS v5.3 PVcalc outputs.totals.fixed.l_tg (data/raw/pvcalc_*_loss14_t30_a0.json) |
| inverter efficiencyFronius Primo GEN24 Plus European efficiency 96.8-97.2%; Huawei SUN2000 residential max efficiency 98.2-98.6%. Already inside PVGIS's 14%. | low: 96.8 high: 98.6 % | range | krannich-solar.com, solar.huawei.com |
| module temp coefficientExample datasheet (Jinko Tiger Neo 54HL4R-B). PVGIS uses its own c-Si model. | -0.29 %/K of Pmax | verified | jinkosolar.eu |
| Degradation | |||
| field medianJordan & Kurtz, Photovoltaic Degradation Rates - An Analytical Review, Prog. Photovolt. 2013 (NREL/JA-5200-51664): median 0.5%/yr, mean 0.8%/yr over ~2000 reported rates. | 0.5 %/yr | verified | osti.gov |
| warranty linearLONGi Hi-MO 6: <1.5% first year, 0.40%/yr; Trina Vertex S+ NEG9R.28: 1% first year, 0.4%/yr; Jinko Tiger Neo: 0.40%/yr over 30 years. Warranty limits, not measured rates. | first year: low: 1 high: 1.5 annual after: 0.4 % | range | solar4ever.com.au, jinkosolar.eu, www-cdn.trinasolar.com |
| defaultUse the field median; let the user change it. | 0.5 %/yr | verified | — |
| Battery | |||
| round trip pv to gridTesla Powerwall 3 datasheet: solar-to-battery-to-grid 89%, solar-to-grid 97% (CEC method). Relative to AC surplus: 0.89/0.97 = 0.918. | 0.89 | verified | nyssf.com |
| battery efficiency htwHTW Berlin Stromspeicher-Inspektion 2025, 22 lithium-ion systems, battery only (excl. power electronics). Mean discharge path BAT2AC 95.7% (21 systems); best charge path PV2BAT 98.2%. | mean: 95.7 low: 87.9 high: 97.8 % | verified | solar.htw-berlin.de |
| round trip rangeIndicative only: extremes come from different systems. Tool default is 0.918 relative to AC surplus. | low: 0.8 high: 0.94 | range | solar.htw-berlin.de |
| byd rteBYD Battery-Box Premium HVS datasheet: round-trip >=96% (battery DC side only). | 0.96 | verified | krannich-solar.com |
| depth of dischargeBYD and Huawei LUNA2000 datasheets state usable energy at 100% DoD. Use the datasheet usable kWh as the input, not nominal. | 1 | verified | krannich-solar.com, ske-solar.com |
| warranty years | 10 | verified | nyssf.com, krannich-solar.com |
| capacity fadeNo independent field fade rate sourced. | none | not sourced | — |
| Roof area | |||
| module area per kwpLONGi LR5-54HTH 440-450 W 1722x1134 mm: 4.44-4.34; Jinko JKM425-445N-54HL4R-B 1762x1134 mm: 4.70-4.49; Trina TSM-NEG9R.28 450-475 W 1762x1134 mm: 4.44-4.21. | low: 4.21 high: 4.7 m2/kWp | derived | solar4ever.com.au, jinkosolar.eu, www-cdn.trinasolar.com |
| flat roof row spacingGeometry only; design criterion 'no inter-row shading at noon on 21 Dec' is a choice, not a rule. Roof area per kWp = module area per kWp / GCR, e.g. 4.44/0.585 = 7.6 m2 at 30 deg. Parapet setbacks, access paths and water tanks are not included. | tilt10: 0.783 tilt20: 0.659 tilt30: 0.585 ground cover ratio (module length / row pitch) | derived | — |
| Installed cost | |||
| installed costNo primary source for Maltese installed prices. User input only; no placeholder default. | none | not sourced | — |
| Reference yields | |||
| yield referenceSensitive to the exact point: PVGIS cells near the coast differ by a few percent. | malta S30: 1633.7 malta flat: 1452 malta E30: 1347.1 gozo S30: 1676.3 malta optimum: tilt: 31 azimuth: 0 annual: 1634 gozo optimum: tilt: 32 azimuth: 3 annual: 1677.3 kWh/kWp/yr, loss 14% | verified | PVGIS v5.3 raw files, see data/SOURCES.md |
14. Open data and how to reproduce it
The two files the calculators load are public:
- pv-yield-malta.json (75 kB): yearly and monthly kWh per kWp at 14% loss, indexed annual[tilt][azimuth] and monthly[tilt][azimuth][month], plus the hourly shares and each point's PVGIS optimum.
- constants.json (21 kB): tariffs, feed-in tariff, grant terms, CO₂ factor, losses, degradation, battery and roof figures, each with value, unit, status, source URL, retrieval date and notes.
To rebuild the yield table yourself, with nothing but Python 3 and no API key:
- For each island and each tilt (0, 5 ... 50, 60, 70, 80, 90) and aspect (-165 to 180 in steps of 15), request PVGIS PVcalc as JSON, for example
https://re.jrc.ec.europa.eu/api/v5_3/PVcalc?lat=35.90&lon=14.51&peakpower=1&loss=14&angle=30&aspect=0&mountingplace=free&pvtechchoice=crystSi&outputformat=json. Pace the requests; PVGIS asks for no more than a few per second. - Take
outputs.totals.fixed.E_yand the twelveoutputs.monthly.fixed[].E_minto a tilt by azimuth grid; aspect -180 is the same as 180. - For the hourly shares, request
seriescalcfor 2023 withpvcalculation=1, tilt 30 and aspects 0, -90 and 90, convert the UTC timestamps to Malta time and divide each hour's mean by the day's total for each month. - Interpolate as in section 1 and compare with fresh PVcalc runs at orientations off the grid.
Our download, build and self-check scripts follow exactly these steps; they are kept with the raw PVGIS outputs in the site's source repository rather than on the public site.