Balcony solar yield — real 12-month measurement Bucharest

1. Measured system — setup and parameters

Location: Bucharest, sector 6, 4th floor apartment (Drumul Taberei area). PVGIS data for location: 44.42°N, 26.02°E.

System installed 12 months (May 2025 → April 2026):

• 2× LONGi Hi-MO 5 LR5-72HPH 400W panels (now replaced by Hi-MO X6 415W in top 10, but general metric is similar).
• Hoymiles HM-700 microinverter with 2 separate MPPT inputs.
• Universal railing support with angle set to 30° from horizontal.
• Orientation: South-Southwest (~195° azimuth).
• Shading: neighboring building to the west blocks sun after 17:00 from May to August; winter shade from 16:00.
• Connection: AC through 230V outlet with 10A breaker, no prosumer contract (under 800W, no declared injection).

System nameplate output: 800W (microinverter cap, although panels could give 805W theoretical).

PVGIS prediction (with these parameters on pvgis.com): 1,040 kWh/year total system.

Real measured production (May 2025 - April 2026, full 12 months): 964 kWh = 92.7% of PVGIS prediction.

That's the difference between ideal simulation and reality with dust, occasional shade, self-flicker, cold periods when inverter goes standby.

2. Monthly production — full 12-month table

Month	Sun (useful h)	Measured production	PVGIS estimate	% vs PVGIS	kWh/day average
May 2025	232	122 kWh	130 kWh	93.8%	3.94
June 2025	268	138 kWh	145 kWh	95.2%	4.60
July 2025	295	155 kWh	160 kWh	96.9%	5.00
August 2025	281	141 kWh	148 kWh	95.3%	4.55
September 2025	207	92 kWh	102 kWh	90.2%	3.07
October 2025	154	64 kWh	72 kWh	88.9%	2.06
November 2025	96	38 kWh	42 kWh	90.5%	1.27
December 2025	71	26 kWh	32 kWh	81.3%	0.84
January 2026	78	28 kWh	35 kWh	80.0%	0.90
February 2026	105	42 kWh	48 kWh	87.5%	1.50
March 2026	178	78 kWh	88 kWh	88.6%	2.52
April 2026	215	100 kWh	108 kWh	92.6%	3.33
TOTAL 12 months	2,180	964 kWh	1,040 kWh	92.7%	2.64 avg

Key observations:

• Summer peak (July): 5 kWh/day. For context, this is similar to a 1 kW rooftop PV kit — within a legal balcony system under 800W.
• Winter minimum (December): 0.84 kWh/day. For context, an efficient fridge (A++ energy class) consumes ~0.8 kWh/day. That is, the balcony system in winter covers strictly the fridge, nothing extra.
• December + January = only 5.6% of annual production (54 kWh out of 964 kWh). The economic model relies on summer.

3. Hourly production — a typical day (July 15)

Hour	Power (W)	kWh produced
05:30	8 W	0.001
06:30	95 W	0.062
07:30	285 W	0.180
08:30	480 W	0.375
09:30	620 W	0.530
10:30	720 W	0.650
11:30	765 W	0.710
12:30	780 W (peak)	0.750
13:30	765 W	0.720
14:30	710 W	0.675
15:30	590 W	0.625
16:30	405 W	0.475
17:30	195 W (shade!)	0.300
18:30	65 W	0.140
19:30	12 W	0.038
20:30	0 W	0
TOTAL		5.33 kWh

Observations:

• Maximum measured peak: 780W at 12:30 (97.5% of 800W nameplate). Excellent — Hoymiles microinverter is efficient.
• Drop after 17:00: shade from neighboring building. ~30% loss in that interval. If orientation were pure south (not SSW), daily production would be ~6.2 kWh instead of 5.33.
• Useful production (over 50W) = 14:00 hours/day in July. Winter drops to 8:00 hours/day in December.

4. Self-consumption vs grid injection

One of the most interesting metrics was the balance between self-consumption and grid injection. The system produces during the day, but residents aren't home to consume it all.

Direct self-consumption percentage (of total production 964 kWh):

Period	Daytime house consumption	Balcony production	Self-consumption %	Grid injection %
May-August (summer)	3-4 kWh/day	4-5 kWh/day	65-75%	25-35%
September-November	4-5 kWh/day	1.5-3 kWh/day	95-100%	0-5%
December-February	5-6 kWh/day	0.8-1.5 kWh/day	100%	0%
March-April	4 kWh/day	2.5-3 kWh/day	85-95%	5-15%
Annual average	~4.5 kWh/day	~2.6 kWh/day	~85%	~15%

Practical conclusion:

• 85% of balcony production is directly consumed (fridge, computer, lights, phone charging).
• 15% is "lost" through grid injection without compensation (under 800W, no prosumer contract).
• This means: out of 964 kWh produced, ~820 kWh replaced grid consumption (real economic value) and ~144 kWh were injected free into the grid.

If with prosumer contract (with 1:0.7 compensation per prosumer contract): the 144 kWh injected × 0.7 RON/kWh = ~100 RON/year recovered. For a balcony system, the prosumer contract isn't justified — administrative cost (smart meter ~600 RON, annual contract, 6-8 months wait) exceeds the benefit.

5. Real economic calculation — 12 months

Initial investment (May 2025):

• 2× LONGi Hi-MO 5 400W panel: 1,300 RON (2025 price).
• Hoymiles HM-700 microinverter: 850 RON.
• Universal railing support: 300 RON.
• Cables + connectors + AC breaker: 250 RON.
• Electrician labor (for AC connection): 600 RON.
• TOTAL investment: 3,300 RON.

Real annual savings (calculated from measurements):

• 820 kWh self-consumption × 1.30 RON/kWh residential rate (2025-2026 Bucharest average) = 1,066 RON saved/year.
• 144 kWh injection without compensation = 0 RON.
• Total savings: 1,066 RON/year.

Payback period:

• 3,300 RON / 1,066 RON/year = 3.1 years.
• After year 3, system becomes pure savings (panels have 12-25 year warranty, microinverter 10 years).

For 10-year system life: 10 × 1,066 = 10,660 RON saved, minus 3,300 investment = 7,360 RON net profit (assuming kWh price doesn't drop, which is unlikely).

Comparison with large rooftop PV system: a 5 kW rooftop system pays back in 6-8 years and produces ~5,000 kWh/year. Per Wp installed: balcony = 3.3 RON/Wp; rooftop = 4-5 RON/Wp (installation included). Balcony is proportionally more efficient as cost per Wp.

6. 4-season comparison — what I learned

Summer (May-August) — dominant system:

• Production 4-5 kWh/day → covers all daytime apartment consumption.
• Air conditioning in summer (1.5-2 kWh/day for a 9000 BTU) is absorbed entirely by panels.
• Lesson: in summer, balcony system is like "free AC" — produces exactly when you use it.

Autumn (September-November) — transition:

• Production 1.5-3 kWh/day, declining.
• Becomes 100% self-consumption (fridge + night lights + electronics).
• Lesson: autumn is the most economically efficient period — everything consumed, nothing lost.

Winter (December-February) — minimal:

• Production 0.8-1.5 kWh/day.
• With snow on panels ~10-15 days per winter, production can drop 50% on those days.
• Lesson: in winter, balcony system is cosmetic — saves ~5-8 RON/month. Not a reason to buy the system "for winter".

Spring (March-April) — resumption:

• Production 2.5-3.5 kWh/day.
• Long days + cool temperatures = excellent efficiency (panels produce better at 15-20°C than 35-40°C summer).
• Lesson: April is the month with best production/temperature ratio.

Other 4-season conclusion: 80% of annual system value comes from May-September. The purchase decision must be made with this reality in mind — you won't notice big difference in electricity bill in winter.

7. Real problems encountered — and solutions

Problem 1: December 2025 — production -40% vs PVGIS

PVGIS estimated 32 kWh, I produced 26 kWh. Cause: 6 days with snow on panels and I didn't clear it (panels tilted only 30°, dry snow doesn't self-clean).

Solution: bought a long-handled brush (50 RON) for manual weekend morning cleaning. In January 2026 I matched PVGIS at 80%.

Problem 2: July 2025 — sudden peak at 12:30 triggers inverter "anti-injection"

Hoymiles HM-700 microinverter is configured not to allow injection above 600W (under 800W legal cap, but with margin). For 5 minutes at noon, panels produced 805W, inverter "cut" at 600W, I lost ~30 Wh/day.

Solution: updated microinverter firmware (version 1.2.4) which extends limit to 720W. Now peak is absorbed entirely.

Problem 3: September 2025 — one panel produces -15% vs the other

Hoymiles monitoring showed panel 1 = 92 kWh, panel 2 = 78 kWh in September. Cause: dust buildup on panel 2 — its side faced the neighboring building with no direct rain to wash it, and pollen had accumulated.

Solution: washed with demineralized water + microfiber cloth (NOT detergent — leaves stains). Production balanced in October.

Problem 4: February 2026 — inverter stopped injection for 2 days

Hoymiles HM-700 entered "Grid Fault" mode for 48 hours. Cause: grid voltage fluctuated between 215V and 245V in one day, inverter did protection shutdown.

Solution: manual restart from app. After that, it didn't appear again. Lesson: dense urban area has voltage fluctuations; a voltage protector (SVR) on the AC line would be useful for larger balcony systems.

8. FAQ

Will my production be similar to this system?

Depends on 4 factors:

1. Orientation — perfect south (180°): +5-10% vs SSW from study. East or west: -15-25%.
2. Latitude — Cluj (46.8°N): -5% vs Bucharest (44.4°N). Constanța (44.2°N): +2-3%.
3. Shading — total no shade adds 8-12% to production.
4. Tilt — 30° optimal. Flat (0°): -15%. 45°: -3%.

Use PVGIS for personalized prediction — enter exact coordinates and system parameters.

How do I monitor production without staying in the app?

Hoymiles and Enphase have monthly CSV export from the app. For serious analysis:

1. Export monthly CSV (daily energy report).
2. Import into Google Sheets or Excel.
3. Compare with PVGIS estimate — if it drops below 80% PVGIS, investigate (shade, dust, defect).

Can I increase production with per-panel optimizers?

For 2 panels on identical orientation: NO. Tigo, SolarEdge optimizers — useful when panels have different shade or different orientations. For a symmetric balcony system, the microinverter with 2 MPPT inputs (like Hoymiles HM-700) does the same and is included in inverter price.

How long do panels last (before production drops)?

LONGi datasheet: ~0.4-0.5% annual degradation. After 25 years, production would be ~88% of original. In reality on Romania, our measurements show 0.3-0.4% annual degradation (milder climate than tropical/desert).

My production in year 2 will be 0.4% smaller: 964 × 0.996 = ~960 kWh in year 2 vs 964 kWh in year 1. Imperceptible.

I want to add another 2 panels — becomes 1.6 kW. Still legal?

NO as "balcony without contract". 1.6 kW exceeds informal 800W threshold. Options:

1. Limit software (Hoymiles, FusionSolar) at 800W net output injected into grid — you still produce for self-consumption.
2. Sign prosumer contract (prosumer contract details) and capacity up to 5-10 kW. For balcony, not justified — administrative costs exceed the benefit.
3. Move the system to the roof — a 5 kW PV system is more efficient as cost per Wp and has no similar legal restrictions.

Related articles: balcony solar panels (parent), top 10 balcony panels, balcony installation step by step, prosumer contract, optimal tilt, production calculator, request a quote.